Patent Application
20240177586
The present invention relates to techniques, including a method, and system, for processing audio, motion, ultra wide band (“UWB”) and frequency modulated continuous wave (“FMCW”) signals using a plurality of antenna array, and other conditions and events. In particular, the invention provides an apparatus using multi-core processors and artificial intelligence processes. Merely by way of examples, various applications can include daily life, and others.
In an example, the present invention provides a system and method for monitoring human activity. The system has a stand alone housing, which has a processing platform, an artificial intelligence module, and a plurality of sensing devices, including rf sensors, audio sensors, and motion sensors, each of which communicates information to the artificial intelligence module for processing.
Various conventional techniques exist for monitoring people within a home or building environment. Such techniques include use of cameras to view a person. Other techniques include a pendant or other sensing device that is placed on the person to monitor his/her movement. Examples include Personal Emergency Response Systems (PERS) devices such as LifeAlert® and Philips® LifeLine—each of which are just panic buttons for seniors to press in case of an emergency. Unfortunately, all of these techniques have limitations. That is, each of these techniques fails to provide a reliable and high quality signal to accurately detect a fall or other life activity of the person being monitored. Many people often forget to wear the pendant or a power source for the pendant runs out. Also, elderly people do not want to look like they are old so often times, elderly people do not wear the pendant.
From the above, it is seen that techniques for identifying and monitoring a person is highly desirable.
According to the present invention, techniques related to a method, and system, for processing audio, UWB, FMCW signals using a plurality of antenna array, and other signals and events are provided. In particular, the invention provides an apparatus using multi-core processors and artificial intelligence processes. Merely by way of examples, various applications can include daily life, and others.
According to one aspect of the invention, a method for a computing system is described. One technique includes beaming with a radar emitter from a central hub, a plurality of radar signals into an interior room and receiving with a radar receiver within the central hub, a plurality of radar reflections in response to the plurality of radar signals. One process includes processing with a processor within the central hub, the plurality of radar reflections to form a plurality of processed radar signals, determining with an AI processor within the central hub, at least one activity from a plurality of activities in response to the plurality of processed radar signals, and determining with a central processing unit within the central hub, a notification action to perform in response to the one or more activities.
According to another aspect of the invention, a computing system is disclosed. A device may include a radar emitter configured to transmit a plurality of radar signals into an interior region, and a radar receiver configured to receive a plurality of radar reflections in response to the plurality of radar signals. An apparatus may include a digital signal processor coupled to the radar emitter and the radar receiver, wherein the processor is configured to determine a plurality of processed radar signals in response to the plurality of radar reflections, and an AI processor coupled to the digital signal processor, wherein the AI processor is configured to determine at least one activity from a plurality of activities in response to the plurality of processed radar signals. A system may include a processing unit coupled to the AI processor, wherein the processing unit is configured to determine a notification action to perform in response to the one or more activities, and a remote communications portion coupled to the processing unit, wherein the remote communications portion is configured to communicate occurrence of the one or more activities to a remote location.
According to yet another aspect of the invention, a method for a system is disclosed. One technique may include beaming with a plurality of radar emitters, a plurality of radar signals into an interior region, and receiving with a plurality of radar receivers, a plurality of radar reflections associated with a first object and a second object in response to the plurality of radar signals. A method may include processing with a digital signal processor, the plurality of radar reflections to determine a plurality of processed radar signals primarily associated with the first object, and determining with a graphical processing unit, at least one activity from a plurality of activities in response to the plurality of processed radar signals. A process may include determining with a central processing unit, a notification action to perform in response to the one or more activities, and performing with the central processing unit within the central hub, the notification action.
The above examples and implementations are not necessarily inclusive or exclusive of each other and may be combined in any manner that is non-conflicting and otherwise possible, whether they be presented in association with a same, or a different, embodiment or example or implementation. The description of one embodiment or implementation is not intended to be limiting with respect to other embodiments and/or implementations. Also, any one or more function, step, operation, or technique described elsewhere in this specification may, in alternative implementations, be combined with any one or more function, step, operation, or technique described in the summary. Thus, the above examples implementations are illustrative, rather than limiting.
According to the present invention, techniques related to a method, and system, for processing UWB and FMCW signals using a plurality of antenna array are provided. In an example, the plurality of antenna array, including a receiving antenna array and a transmitting antenna array configured to capture and transmit signals in an omni-directional manner. In particular, the invention provides an apparatus using multi-core processors and artificial intelligence processes. Merely by way of examples, various applications can include daily life, and others.
In an example, multiple aspects of antenna design can improve the performance of the activities of daily life (“ADL”) system. For example in scanning mode the present technique continuously looks for moving human targets (or user) to extract ADL or fall. Since these can happen anywhere in the spatial region of a home, the present system has antennas that have wide field of view. Once the human target is identified, the technique focuses signals coming only from that particular target and attenuate returns from all other targets. This can be done by first estimating location of the target from our technique using wide field of view antennas and then focusing RF energy on the specific target of interest once it has been identified. In an example, the technique can either electronically switch a different antenna that has narrow field of view or could use beam forming techniques to simultaneously transmit waves from multiple transmit antenna and control their phase such that the RF energy constructively builds around the target of interest where as it destructively cancels everywhere else. This return will be much cleaner and can boost the performance of our ADL+fall+vital sign sensors.
In another example considers the layout of the antennas itself. In an example, the technique places transmit and receive antennas in various different physical configurations (ULA, circular, square, etc.), that can help us establish the direction from which the radar signal returns, by comparing phases of the same radar signal at different receiving antennas. The configurations can play a role because different configurations enable direction of arrival measurement from different dimensions. For example, when the human target falls the vertical angle of arrival changes from top to bottom, therefore a vertical ULA is better suited to capture that information. Likewise during walking horizontal angle of arrival of the signal varies more therefore it makes sense to use horizontal ULA is more sensitive and therefor can provide additional information for our algorithm. Of course, there can be other variations, modifications, and alternatives.
In an example, the wireless RF unit can be either pulsed doppler radar or frequency modulated continuous wave (FMCW) or continuous wave doppler (CW). In an example, on the transmit side it will have standard RF units like VCO, PLL, among others. On the receive side it can have matched filter, LNA, mixer, and other elements. The multiple antennas can be either driven by a single transmit/receive chain by sharing it in time or have one each chain for each of the antennas.
In an example, waveform pattern generator generates control signals that define the type of radar signal that is generated by the radar RF unit. For example for FMCW, it can generate triangular wave of specific slope and period, which will linearly sweep the frequency of the RF unit according to this parameter. For a pulsed doppler radar, the technique will hold generate pulse of specific width and period, which will modulate the RF output accordingly.
In an example, the gain and filter stage filters the radar returns to remove any unwanted signals and then amplifies the remaining signal with different techniques. For example, the present artificial intelligence or AI technique can determine what target is desirably tracked and provide feedback to the AI technique, that will filter out radar return from any and all other signals except for the signal that is desirably tracked. If human target is moving the return signal will be fluctuating, in that case, the technique applies automatic gain control (AGC) to find the optimal gain, so that entire dynamic range of ADC in the subsequent stage is satisfied. In an example, the return signal is converted to digital samples by analog-to-digital converters (ADC), among other frontend elements.
In an example, the present technique is provided to identify various activities in home using non-wearable. In an example, the technique is at least privacy intrusive as possible, and will use sensors that are less intrusive. Examples of sensors can include, without limitation, a wireless backscatter (e.g., radar, WiFi), audio (e.g., microphone array, speaker array), video (e.g., PTZ mounted, stereo), pressure mats, infrared, temperature, ultraviolet, humidity, pressure, smoke, any combination thereof, and others.
In an example, the technique can use wireless backscattering to measure motion of human, a location, and an environmental state, such as door opening/closing, or other environmental condition. In an example, the wireless backscattering can also be used to measure a vital sign, such as a heart rate and respiration rate, among others. In an example, the wireless techniques can work in non-line of sight, and is non intrusive compared to camera or microphone, or others. In an example, the technique can use radar\backscatter sensor for two purposes (1) to find the location of an action; and (2) sense different activities associated with the action. Of course, there can be other variations, modifications, and alternatives.
In an example, the present technique and system includes a radar system that operates on multiple frequency bands, such as below 10 GHz, around 24 GHz, 60 GHz, 77-81 GHz, among others. In an example, different frequency interacts differently with various objects in our environment. In an example, available signal bandwidth and permissible signal power are also regulated differently at different frequency bands. In an example, the present techniques optimally combine reflections coming from a reflector from multiple frequency bands to achieve large coverage, and/or improve accuracy. Of course, there can be other variations, modifications, and alternatives.
In an example, each radar is working at a particular frequency band will be using multiple transmit and receive antennas, as shown. In an example, using these multiple transmitters, the technique can perform transmit beam forming to concentrate radar signal on a particular target. In an example, the technique uses multiple receivers to collect reflected signals coming from various reflectors (e.g., human body, walls). After further processing this will allow us to find the direction of the reflector with respect to the radar. In an example, the technique also uses multiple transmitter and receiver to form virtual array, this will allow emulate the radar array with large element by using small number of transmitter and receiver chains. The main benefit is to improve the angle resolution without using a large array, saving space and component cost. In an example, different antenna array configurations to improve coverage (using beam forming) or add 3D localization capability (using 2-D array) are included.
In an example using standard radar signal modulation techniques, such as FMCW/UWB, on MIMO radar, the technique will first separate signals coming from different range and angle. The technique will then identify static reflectors, such as chairs, walls, or other features, from moving ones, such as human targets, pets, or the like. For moving objects that are tracked, the technique will further process signals for each of the reflectors. As an example, the technique will use different techniques to extract raw motion data (e.g., like spectrogram). In an example, the technique will apply various filtering process to extract periodic signals generated by vital signs, such as heart rate, respiration rate, among others. In an example, both the raw motion data and extracted vital signs will be passed to a downstream process, where they are combined with data from other sensors, such as radar outputs operating at different frequency or completely different sensors to extract higher insights about the environment. Of course, there can be other variations, modifications, and alternatives.
In an example, the present technique uses a sensor array that has a multiple microphone array. In an example, these microphones will be use to ascertain the direction of arrival of any audio signal in the environment. In an example, the microphone in conjunction with other sensors, such as radar, will be vital in performing two tasks: 1st it will augment radar signal to identify various activities (walking produces a different sound than sitting), if the target is watching TV it is much easier to ascertain it with audio signal; and 2nd in case of emergency like fall, the technique can use the radar signal to identify the location of the fall and then beam form microphone array towards that location, so that any audio signal produced by the target can be captured. Of course, there can be other variations, modifications, and alternatives.
In addition to a radar sensor, which is consider as active sensors the present sensor system (e.g., box, boxes) will also have additional passive sensors that captures the sound, chemical signature, environmental conditions. Each of these of the sensors captures different context about the home that the human being tracking is living in or occupying. In an example, the UV sensor can monitor how often the sunlight comes in the room. In an example, light sensors determine a lighting condition of the human's home or living area.
In an example, a microphone array can have many functions, such as use to sense sound in the room, to figure out how long the human has spent watching TV, or how many time they went to bathroom by listening to the sound of toilet flushing or other audio signature. In an example, the present technique can use creative solutions where it can use the active sensor to find the location of the person and then tune the microphone array to enhance the sound coming from that location only, among other features. In an example, the technique can call the sensors that are derived from the hardware sensors using specific algorithms as software sensors or soft sensors. So the same hardware sensors can be used for many different applications by creating different software sensors. Here the software sensors can combine signals from one or more sensors and then apply sensor fusion and AI techniques to generate the desired output. Of course, there can be other variations, modifications, and alternatives.
In example, radar sensors can determine information about a human's location within a home, like if they are in kitchen area, or other. In an example, when the human target turns on the microphone oven, it generates specific RF signature that can be tracked. In an example, the technique can combine this information to infer if the human target walked to the kitchen and turned on the microphone. Likewise, when the human target prepares food in kitchen he/she can make lot of specific noise like utensils clattering, chopping, or other audio signature. So if a human target goes to kitchen spends sometime time in the kitchen, and the present microphone pick these sounds, the technique can infer that food is cooking or other activity.
In an example, toileting frequency can be a very valuable indication of ones wellness. The present technique can track if a human went to the bathroom using the radar or other sensing techniques. In an example, additionally, the technique can pick sound signature of toilet flushing. In an example, the technique combines these two pieces of information, which can be correlated to toileting frequency. In an example, similarly, bathing is a unique activity that requires 4-5 minutes of specific movements. By learning those patterns, the technique can figure out ones bathing routines.
In an example, different sensors are triggered by different motion of a human target. In an example, radar can detect human fall by looking at micro doppler patterns generating by different part of the target during falls. In an example, the technique can also simultaneously hear a fall from microphone arrays and vibration sensors. In an example, the technique can also detect how pace of movement changes for an individual over a long duration by monitoring the location information provided by radar or other sensing technique. In an example, likewise, the technique can gather unstable transfers by analyzing the gait of the target. In an example, the technique can find front door loitering by analyzing the radar signal pattern. In an example, the technique can figure out immobility by analyzing the radar return. In this case, the technique can figure out the target's presence by analyzing the target's vital signs, such as respiration rate or heart rate or by keeping track of the bread crumb of the target's location trace.
In any and all of the above cases, the technique can also learn about the exact environmental condition that triggered a particular state. For example, the technique can figure out whether a human target was immobile because the target was watching TV or a video for long duration or the target was simply spending a lot of time in their bed. And these can be used to devise incentives to change the target's behavioral pattern for better living.
In an example, the technique can estimate vital signs of a person by sensing the vibration of the target's body in response to the breathing or heart beat, each of the actions results in tiny phase change in the radar return signals, which can be detected. In an example, the technique will use several signal processing techniques to extract them. Of course, there can be other variations, modifications, and alternatives.
In an example, different frequency radio wave interact with environment differently. Also phase change due to vital signs (HR,RR) differs by frequency, for example phase change for a 77 GHz radar is much higher than for a 10 GHz radar. Thus 77 GHz is more appropriate for estimating heart-beat more accurately. But higher frequency typically attenuates much more rapidly with distance. Therefore, lower frequency radar can have much larger range. By using multi-frequency radar in the present technique can perform these vital trade-offs.
In an example, the present radar sensors can detect motions that are generated during sleep, such as tossing and turning. In an example, radar sensors can also sense vital signs like respiration rate and heart rate as described earlier. In an example, now combining the pattern of toss and turn and different breathing and heart beat pattern, the technique can effectively monitor the target's sleep. Additionally, the technique can now combine results from passive sensors, such as a thermometer, UV, photo diode, among others, to find correlation between certain sleep pattern and the environmental conditions. In an example, the technique can also use the sleep monitor soft sensor to learn about day/night reversal of sleep, and the associated environmental condition by looking at different passive sensors. In an example, the techniques can be valuable in providing feedback to improve the human target's sleep. For example, the technique can determine or learn that certain environmental condition results in better sleep and prescribe that to improve future sleep.
In an example, the technique can repurpose many of the sensors described before for security applications. For a security application, the technique determines where one or more person is located, which can be detected using a presence detection sensor that is build on top of radar signals. In an example, the technique can eliminate one or many false positive triggered by traditional security systems. For example, is a window is suddenly opened by a wind the technique (and system) will look at presence of human in the vicinity before triggering the alarm. Likewise, combination of vital signs, movement patterns, among others, can be used a biometric to identify any human target. If an unknown human target is detected in the vicinity at certain time of the day, the technique can trigger an alarm or alert.
In an example, any one of the above sensing techniques can be combined, separated, or integrated. In an example, n addition to radar and audio sensors, other sensors can be provided in the sensor array. Of course, there can be other variations, modifications, and alternatives.
In an example, the hub includes various sensing devices. The sensing devices, include, among others, a radar, a WiFi, a Bluetooth, a Zigbee sniffer, a microphone and speakers, a smoke detector, a temperature detector, a humidity detector, a UV detector, a pressure detector, MEMS (e.g., accelerometer, gyroscope, and compass), a UWB sensors (for finding locations of all the deployed elements relative to each other), among others. In an example, the hub is a gateway to internet via WiFi, GSM, Ethernet, landline, or other technique. The hub also connects to other units (Mini Node/Mobile Node) via Bluetooth, WiFi, Zigbee, UWB and coordinates them with each other. In an example, certain data processing, such as noise removal, feature extraction to reduce amount of data uploaded to cloud is included. In an example, the hub alone can be sufficient to cover a small living space. In an example, the hub is deployed as a single device somewhere in a desirable location (e.g., middle of the living space) so that it has good connectivity to all other units. An example of such deployment is provided in the Figure below.
In an example, as shown in
In an example, the housing has a height characterizing the housing from a bottom region to a top region. In an example, a plurality of levels 803 are within the housing numbered from 1 to N, wherein N is an integer greater than two, but can be three, four, five, six, seven, and others.
As shown, various elements are included. A speaker device 809 configured within the housing and over the bottom region, as shown. The hub device also has a compute module 811 comprising a processing device (e.g., microprocessor) over the speaker device. The device has an artificial intelligence module configured over the compute module, a ultra-wide band (“UWB”) module 813 comprising an antenna array configured over the artificial intelligence module, and a frequency modulated continuous wave (“FMCW”) module 815 with an antenna array configured over the UWC module. In an example, the FMCW module being configured to process electromagnetic radiation in a frequency range of 24 GHz to 24.25 GHz. In an example, the FMCW module outputs an FMCW signal using a transmitter, and receives back scattered signals using a receiver, such as a receiver antenna. The device has an audio module configured over the FMWC module and an inertial measurement unit (“IMU”) module configured over the FMCW module. In an example, the audio module comprises a microphone array for detecting energy in a frequency range of sound for communication and for detecting a sound energy. In an example, the IMU module comprises at least one motion detection sensor consisting of one of a gyroscope, an accelerometer, a magnetic sensor, or other motion sensor, and combinations thereof.
As shown, the speaker device, the compute module, the artificial intelligence module, the UWB module, the FMCW module, the audio module, and the IMU module are arranged in a stacked configuration and configured, respectively, in the plurality of levels numbered from 1 to N. In an example, the speaker device comprises an audio output configured to be included in the housing. As shown, the speaker device is spatially configured to output energy within a 360 degree range from a midpoint of the device.
In an example, the compute module comprises a microprocessor based unit coupled to a bus. In an example, the compute module comprises a signal processing core, a micro processor core for an operating system, a synchronizing processing core configured to time stamp, and synchronize incoming information from each of the FMCW module, IMU module, and UWB module.
In an example, the device further comprises a real time processing unit configured to control the FMCW switch or the UWB switch or other switch requiring a real time switching operation of less than ½ milliseconds of receiving feedback from a plurality of sensors.
In an example, the device has a graphical processing unit configured to process information from the artificial intelligence module. In an example, the artificial intelligence module comprises an artificial intelligence inference accelerator configured to apply a trained module using a neural net based process. In an example, the neural net based process comprises a plurality of nodes numbered form 1 through N. Further details of the UWB module can be found throughout the specification and more particularly below.
In an example, the device has a triangular configuration comprising a first antenna array, a second antenna array, and a third antenna array included in the at least three antenna arrays. The three arrays provide a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured from a vertical plane normal to the horizontal plane. As previously noted, the three arrays are enclosed in a housing that provides mechanical support. In an example, each of the sensor arrays is provided on a substrate member to be configured in the triangular configuration. The substrate member has a face arranged in a normal manner in a direction to each of the support members.
In an example, the UWB module can operate at a center frequency of 7.29 GHz and a bandwidth of ˜1.5 GHz with multiple antenna arrays to achieve the FCC/ETSI compliance standard. In an example, the module has a combined horizontal field-of-view of 360 degrees about a center point of the module. In an example, the module has a range greater than 10 meters, but can be shorter and longer. In an example, the module is configured to achieve a transmission and a receive rate of frames per second (FPS) equal to or greater than 330 per Tx-Rx. In an example, the module has a combined horizontal field of view of 360 degrees achieved using three (3) antenna arrays, each of which covering 120 degrees. In an example, each antenna array comprises of 1-TX and 4-RX. Each antenna array is configured to complete the acquisition of a frame within 1 millisecond or less. Accordingly, a total of three (3) milliseconds covers all three (3) sectors, achieving a frame rate of 330 fps per sector (per Tx-Rx) in an example. In an example, the module has programmability of various parameters similar to Novelda X4M03 module. In an example, the module is a hybrid architecture that has four by four radar integrated circuit devices in MIMO configuration that switches between the three antenna arrays. The configuration is capable of simultaneous capturing of all four Rx frames in an antenna array. Further details of the present UWB module is provided throughout the present specification and more particularly below.
In an example, the basic components of the X4 SoC are a transmitter, a receiver, and related control circuits. The system is controlled by a system controller and is configurable through a 4(6)-wire serial peripheral interface (SPI). In an example, the X4 receive path (RX) consists of a low noise amplifier (LNA), a digital-to-analog converter (DAC), 1536 parallel digital integrators as well as an output memory buffer, accessible through the SPI. The RX is tightly integrated with the transmitter (TX) and is designed for coherent integration of the received energy. The X4 transmit path (TX) consists of a pulse generator capable of generating pulses at a rate of up to 60.75 MHz. The output frequency and bandwidth are designed to fit worldwide regulatory requirements. The radar transceiver is able to operate completely autonomously and can be programmed to capture data at predefined intervals and then alert or wake up a host MCU or DSP through dedicated interrupt pins. A power management unit controls the on-chip voltage regulators and enables low-power applications to use efficient duty cycling by powering down parts of the circuit when they are not needed. The system can be configured to consume less than 1 mW in idle mode when all analog front end components are turned off. As shown, each of the four X4 SoCs is coupled in parallel to a switch.
In an example, the switch 1109 is coupled to each antenna array as shown. In an example, the switch can be one listed under HMC241/HMC7992/ADRF5040 SP4T RF Switches of Analog Devices, Inc. The switches are non-reflective RF switches from DC to 12 GHz for 4G cellular, milcom, and radio applications. Examples of HMC241, HMC7992, and ADF5040 are radio frequency (RF) nonreflective/absorptive single pull, quad throw (SP4T) switches that can interface with 3.3 V, TTL, LVTTL, CMOS, and LVCMOS logic. The switches operate from DC to 12 GHz frequency range. The HMC241 is a GaAs MMIC RF switch that operates in the DC to 4 GHz range. The switch takes a single supply at +5 V. The HMC7992 has a 100 MHz to 6 GHz frequency range. The ESD rating is for this switch 2 kV (HBM) class 2. The HMC7992 takes a single voltage supply from ±3.3 V to +5 V. The ADRF5040 comes in a small 4 mm×4 mm LFCSP package and requires a dual ±3.3 V supply. The switch operates in the 9 kHz to 12 GHz range. The ADRF5040 has the added benefit of being 4 kV (HBM) ESD rating. HMC241, HMC7992, and ADF5040 are ideal for 4G cellular infrastructure such as base stations and repeaters as well as military communications and industrial test and measurement applications. Of course, there can be other variations, modifications, and alternatives.
In an example, the UWC module comprises a switch configured between a plurality of UWC transceivers. The switch is configured to select one of the three antenna arrays to sense the back scatters while the other two antenna arrays are turned off. In an example, the switch is an rf switch such as the one listed under part number ADRF-5040 manufactured by Analog Devices, Inc. In an example, the UWC module also has a controller configured to control the switch and the three antenna array. In an example, the controller cycles through a predetermined process to decide which one of the three antenna array to activate while the other two antenna arrays are turned off.
In an example, the at least three antenna array are configured to sense electromagnetic energy ranging from 6 to 8 GHz in frequency. As noted, the sensing apparatus is spatially positioned within a center of a geographic location of a room to detect movement of human user.
In an example, the present invention provides a method processing an electromagnetic signal generated from an ultra wide band rf signal to detect an activity of a human user. Referring to
In an example, the present invention provides a method processing an electromagnetic signal generated from an ultra wide band rf signal to detect an activity of a human user. In an example, the method includes generating a base band outgoing UWC signal. The method also includes receiving the base band outgoing UWC signal at a switch device and directing the outgoing UWC signal using the switch device to one of three antenna arrays configured in a triangular configuration to transmit the outgoing UWC signal from spatial location of a zero degree location in relation to a mid point of the device through a 360 degrees visibility range where each antenna array is configured to sense a 120 degree range in a horizontal plane. Each of the antenna array is configured to sense and transmit at least an 80 degree visibility range as measured from a vertical plane that is normal to the horizontal plane.
In an example, each of the three antenna arrays has a support member, e.g., board, printed circuit board. In an example, each array has a plurality of transmitting antenna spatially configured on a first portion of the support member, a transmitting integrated circuit coupled to each of the plurality of transmitting antenna and configured to transmit the outgoing UWC signal, a plurality of receiving antenna spatially configured on second portion of the support member, and a receiving integrated circuit coupled to each of the plurality of receiving antenna and configured to receive an incoming UWB signal and configured to convert the UWC signal into a base band signal. In an example, the method includes receiving a back scattered electromagnetic signal caused by an activity of a human user redirecting the outgoing UWB signal.
The apparatus of claim 11 wherein the UWB module comprises a micro controller unit coupled to a memory resource, and a clock circuit, the micro controller unit being configured with a universal serial bus interface coupled to the compute module; wherein the compute module is configured with the artificial intelligence module to process information from the back scattered electro magnetic signal from the base band signal to detect the activity of the human entity.
In an example, the support member comprises a major plane positioned normal to a direction of gravity.
In an example, the antenna array comprises at least three antenna array spatially arranged in a triangular configuration comprising a first antenna array, a second antenna array, and a third antenna array included in the at least three antenna arrays to provide a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured from a vertical plane normal to the horizontal plane. In an example, the antenna array comprises at least three antenna array spatially arranged in a triangular configuration comprising a first antenna array, a second antenna array, and a third antenna array included in the at least three antenna arrays to provide a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured from a vertical plane normal to the horizontal plane, and further comprising a controller configured to control a switch coupled with each of the three antenna array, the controller cycles through a predetermined process to decide which one of the three antenna array to activate while the other two antenna arrays are turned off.
In an example, each antenna array comprises 1-TX and 4-RX.
In an example, the system has a switch device coupled between each of the antenna array and four receive lanes each of which is coupled to the receiving integrated circuit device, one transmit lane coupled to a transmitting integrated circuit device, and a micro controller unit coupled to a bus coupled to the receiving integrated circuit device and the transmitting integrated circuit device, the micro controller unit coupled to a memory resource configured with the micro controller to store raw data from information derived from four receive lanes, the micro controller unit being coupled to a clock.
In an example, each antenna array comprises 1 TX and four RX. In an example, the system has a switch device coupled between each of the three antenna arrays and four receive lanes each of which is coupled to the receiving integrated circuit device, one transmit lane coupled to a transmitting integrated circuit device, and a micro controller unit coupled to a bus coupled to the receiving integrated circuit device and the transmitting integrated circuit device, the micro controller unit coupled to a memory resource configured with the micro controller to store raw data from information derived from four receive lanes, the micro controller unit being coupled to a clock.
In an example, the present techniques include a method, apparatus, and device for processing signals. As shown 1400 in
Referring now to
As shown in the Table in
In an example,
In an example,
In an example, the system has a processing and acquisition module 1807. The module has a quad channel AFE RX listed under ADAR7251 by Analog Devices, and a FMCW generator listed under ADF4159. The module is coupled to and communicates with a 12 channel—3:1 demux switches 1805 listed under TS3DV621 by Texas Instruments Incorporated. The system has a microcontroller such as a Cypress Microcontroller listed under part number CYYSB301X, which is coupled to a memory resource, such as a 2 GB SDRAM. The system has a SPI Interface control between RF module and microcontroller. A USB interface is coupled to TCP 1809. Of course, there can be other variations, modifications, and alternatives. Further details can be found in a more detailed diagram 1850 of
In an example on a transmit lane 1851 referring to
In an example on a receive lane 1853, four FMCW signals are received from four RX antenna. The four signals are received in parallel, and fed to and processed in the Rf integrated circuit to output corresponding four base band analog signals, each of which is fed to the switch. The switch allows signals from one of the three antenna array to be transferred to corresponding analog to digital converters, each of which are in parallel. Each analog to digital converter is coupled to the microcontroller. Each analog to digital converter configures incoming base band signal into digital, which is fed to the microcontroller. Of course, there can be other variations, modifications, and alternatives.
In an example,
In an example, each antenna array has two (2) TX in an antenna array that are spaced by lambda apart horizontally and lambda over two vertically to form a virtual 2D array with the 4 RX 2003. In an example, the present virtual antenna mapping is provided to achieve the goal of power balancing the physical channels across the multiple physical antennas especially when multiple input multiple output is deployed in the downlink. In an example, virtual antenna mapping gives an illusion that there are actually lesser antennas at the base station than it actually has. The unbalanced balanced power across two transmits paths are transformed into balanced power at physical antenna ports by virtual antenna mapping. This is achieved using phase and amplitude coefficients. Thus both the power amplifiers are optimally used even for signals transmitted on the first antenna. Of course, there can be other variations, modifications, and alternatives.
In an example, use of higher power with FMCW can be used to capture more granular features, such as breathing, heart rate, and other small scale features. In an example, lower power and UWB is desirable for more gross features, which has lower frequency. Lower frequency can also penetrate walls, and other physical features.
In an example, the present invention provides an FMCW sensor apparatus. The apparatus has at least three transceiver modules. Each of the transceiver modules has an antenna array to be configured to sense a back scatter of electromagnetic energy from spatial location of a zero degree location in relation to a mid point of the device through a 360 degrees range where each antenna array is configured to sense a 120 degree range. In an example, each of the antenna array has a support member, a plurality of receiving antenna, a receiver integrated circuit coupled to the receiving antenna and configured to receive an incoming FMCW signal and covert the incoming FMCW signal into a base band signal, and a plurality of transmitting antenna. Each antenna array has a transmitter integrated circuit coupled to the transmitting antenna to transmit an outgoing FMCW signal. The apparatus has a virtual antenna array configured from the plurality of receiving antenna and the plurality of transmitting antenna to form a larger spatial region using the virtual antenna array, than a physical spatial region of the plurality of receiving antenna. In an example, the apparatus has a triangular configuration comprising a first antenna array, a second antenna array, and a third antenna array included in the at least three antenna arrays to provide a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured from a vertical plane normal to the horizontal plane. The apparatus has a master control board coupled to each of the support members, and configured in a normal directional manner with reference to each of the support members. The apparatus has a housing enclosing the at least three transceiver modules.
In an example, the FMCW sensor apparatus comprises a switch configured between a plurality of FMCW transceivers, such that the switch is configured to select one of the three antenna arrays to sense the back scatters while the other two antenna arrays are turned off. In an example, the antenna array is configured to process electromagnetic radiation in a frequency range of 24 GHz to 24.25 GHz.
In an example, apparatus has a controller configured to control the switch and the three antenna array. In an example, the controller cycles through a predetermined process to decide which one of the three antenna array to activate while the other two antenna arrays are turned off. In an example, the three antenna array are configured to sense electromagnetic energy in a 24 GHz to 24.25 GHz frequency band. In an example, the sensing apparatus is spatially positioned within a center of a geographic location of a room to detect movement of human user. In an example, each of the sensor arrays is provided on a substrate member to be configured in the triangular configuration.
In an example, the apparatus has a housing. The housing has a maximum length of six to twenty four inches and width of no longer than six inches. In an example, the housing has sufficient structural strength to stand upright and protect an interior region within the housing.
In an example, the apparatus has a height characterizing the housing from a bottom region to a top region, a plurality of levels within the housing numbered from 1 to N, and a speaker device configured within the housing and over the bottom region. In an example, the apparatus has a compute module comprising a processing device over the speaker device, an artificial intelligence module configured over the compute module, a ultra-wide band (“UWB”) module comprising an antenna array configured over the artificial intelligence module, and an audio module configured over the FMWC module. The apparatus has an inertial measurement unit (“IMU”) module configured over the FMCW module.
In an example, the speaker device, the compute module, the artificial intelligence module, the UWB module, the FMCW module, the audio module, and the IMU module are arranged in a stacked configuration and configured, respectively, in the plurality of levels numbered from 1 to N.
In an example, the speaker device comprises an audio output configured to be included in the housing, the speaker device being configured to output energy within a 360 degree range from a midpoint of the device.
In an example, the compute module comprises a microprocessor based unit coupled to a bus. In example, the compute module comprises a signal processing core, a micro processor core for an operating system, a synchronizing processing core configured to time stamp, and synchronize incoming information from each of the FMCW module, IMU module, and UWB module.
In an example, the apparatus has a real time processing unit configured to control the FMCW switch or the UWB switch or other switch requiring a real time switching operation of less than ½ milliseconds of receiving feedback from a plurality of sensors. In an example, the apparatus has a graphical processing unit configured to process information from the artificial intelligence module.
In an example, the artificial intelligence module comprises an artificial intelligence inference accelerator configured to apply a trained module using a neural net based process, the neural net based process comprising a plurality of nodes numbered form 1 through N.
In an example, the FMCW module comprises at least three antenna arrays to be configured to sense a back scatter of electromagnetic energy from spatial location of a zero degree location in relation to a mid point of the device through a 360 degrees range where each antenna array is configured to sense a 120 degree range.
In an example, each of the antenna arrays comprises a FMCW transceiver and a switch configured between each of the FMCW transceiver and a controller, such that the switch is configured to select one of the three antenna arrays and the FMWC transceiver to sense the back scatters while the other two antenna arrays are turned off, and further comprising a serial interface.
In an example, the audio module comprises a micro phone array for detecting energy in a frequency range of sound for communication and for detecting a sound energy.
In an example, the UMU module comprises a support substrate, an electrical interface provided on the support structure, an accelerometer coupled to the electrical interface, a gyroscope coupled to the electrical interface, a compass coupled to the electrical interface, a UV detector configured to detect ultraviolet radiation coupled to the interface, a pressure sensor coupled to the interface, and an environmental gas detector configured and coupled to the interface to detect a chemical entity.
In an example, the present invention provides an apparatus for processing activities of a human user. The apparatus has an audio module and a compute module coupled to the audio module. The apparatus has a transceiver module coupled to the compute module. In an example, the transceiver module has an antenna array to be configured to sense a back scatter of electromagnetic energy in a frequency range of 24 GHz to 24.25 GHz from spatial location of a zero degree location in relation to a mid point of the device through a 360 degrees range where each antenna array is configured to sense a 120 degree range.
In an example, the antenna array comprises a support member, a plurality of receiving antenna, a receiver integrated circuit coupled to the receiving antenna and configured to receive an incoming frequency modulated continuous wave (FMCW) signal and covert the incoming FMCW signal into a base band signal, a plurality of transmitting antenna, a transmitter integrated circuit coupled to the transmitting antenna to transmit an outgoing FMCW signal.
In an example, the apparatus has a virtual antenna array configured from the plurality of receiving antenna and the plurality of transmitting antenna to form a larger spatial region using the virtual antenna array, than a physical spatial region of the plurality of receiving antenna. In an example the apparatus has a master control board coupled to the support member, and configured in a normal directional manner with reference to the support member and a housing enclosing the transceiver modules, the compute module, and the audio module.
In an example, the present invention has methods using the apparatus, device, and systems. In an example, the method is for processing signals from human activities. The method includes generating an rf signal using a transceiver module coupled to a compute module and emitting the rf signal using one of three antenna array and sensing using one of the three antenna array configured from spatial location of a zero degree location in relation to a mid point of the three antenna array through a 360 degrees range where each antenna array is configured to sense a 120 degree range to capture a back scatter of electromagnetic energy in a frequency range of 24 GHz to 24.25 GHz associated with a human activity.
In an example, the present invention provides an alternative radio frequency (RF) sensing apparatus. The apparatus has an ultra wide band (UWB) module comprising at least three ultra wide band (UWB) antenna arrays configured in a triangular arrangement to sense a back scatter of electromagnetic energy from a spatial location such that the triangular arrangement allows for sensing from a zero degree location in relation to a mid point of the triangular arrangement through a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured from a vertical plane that is normal to the horizontal plane where each UWB antenna array is configured to sense at least a 120 degree range.
In an example, the apparatus has a frequency modulated continuous wave module comprising at least three frequency modulated continuous wave (FMCW) transceiver modules. Each of the FMCW transceiver modules has a FMCW antenna array. In an example, the three FMCW transceiver modules are configured in a triangular arrangement to sense a back scatter of electromagnetic energy from spatial location such that the triangular arrangement allows for sensing from a zero degree location in relation to a mid point of the triangular arrangement through a 360 degree visibility range as measured from a horizontal plane, and a 80 degree visibility range as measured form a vertical plane that is normal to the horizontal plane where each FMCW antenna array is configured to sense at least a 120 degree range.
In an example, the system has multiple interfaces. A USB 3.0 interface communicates to the FMCW module. The I2S interface communicates to the audio module. A USB 2.0 interface communicates to the UWB module. Another USB 2.0 interface communicates to a user interface, such as a keyboard and a mouse. Other types of serial interfaces can also be included. The system also has an RJ-45 and Ethernet interface, a Wi-Fi and Blue Tooth interface, a cellular interface, such as LTE, among others. The system has a global positioning sensor interface. The system has a power and clock module for power and clocking functions. The system has an inertial measurement unit connector and module. The system has multiple PCIE connector interfaces, one of which is coupled to a Wi-Fi sensor device. Other features include dynamic random access memory interface, embedded multi-media card connection and module, a solid disk drive connector, and a serial advanced technology attachment connector, among others.
An example of the processing system can be a single integrated circuit chip manufactured by Texas Instruments Incorporated sold as AM572x Sitara Arm applications processors. In a datasheet by for the Sitara Arm by Texas Instruments, “AM572x devices bring high processing performance through the maximum flexibility of a fully integrated mixed processor solution. The devices also combine programmable video processing with a highly integrated peripheral set. Cryptographic acceleration is available in every AM572x device. Programmability is provided by dual-core Arm Cortex-A15 RISC CPUs with Neon™ extension, and two TI C66x VLIW floating-point DSP cores. The Arm allows developers to keep control functions separate from other algorithms programmed on the DSPs and coprocessors, thus reducing the complexity of the system software. Additionally, TI provides a complete set of development tools for the Arm and C66x DSP, including C compilers, a DSP assembly optimizer to simplify programming and scheduling, and a debugging interface for visibility into source code execution.”
The processing system is coupled to a energy source, including a battery and a plug connection. The system also has a graphical processing module or artificial intelligence module for performing processing functions from data received from the interfaces. An example of the processing unit is one sold under the Movidius™ brand by Intel Corporation.
In an example, Movidius provides low-power vision processing solutions, which include the Myriad 2 family of vision processing units (VPUs) plus a comprehensive Myriad Development Kit (MDK), a reference hardware EVM and optional Machine Vision Application Packages. In an example, The Myriad 2 MA2x5x family of system-on-a-chip (SoC) devices offers significant computation performance and image processing capability with a low-power footprint. The Myriad 2 lineup includes the following product configurations: MA2150: 1 Gbit DDR MA2155: 1 Gbit DDR and secure boot MA2450: 4 Gbit DDR MA2455: 4 Gbit DDR and secure boot.
In an example, the Myriad 2 VPUs offer TeraFLOPS (trillions of floating-point operations per second) of performance within a nominal 1 Watt power envelope. The Myriad 2 architecture includes enough performance to support multiple cameras with flexible image signal processing pipelines for each camera, and software programmable vision processing with fixed- and floating-point datatypes supported. A robust overall dataflow design ensures mitigation of processing bottlenecks.
In an example, Myriad 2 MA2x5x incorporates an innovative approach to combine image signal processing with vision processing. A set of imaging/vision hardware accelerators supports a world-class ISP pipeline without any roundtrips to memory; at the same time they are repurposed to accelerate developers' vision processing algorithms in conjunction with a set of special purpose VLIW vision processor cores. In an example, a vision processor can be configured to process RF signals for filtering and/or categorizing specific signals from the RF signals. In an example, the digital signal processor can also be configured to filter out undesired signals before processing a designed signal. All processing elements are tied together with a multi-ported memory that enables implementation of demanding applications with high efficiency.
In an example, the artificial intelligence module can process data or information by a variety of artificial intelligence techniques. As an example, machine learning uses on one or more computer algorithms that improve automatically through experience. In an example, the technique can use unsupervised learning to find patterns in a stream of input, without requiring a human to label the inputs first. Alternatively, the technique can use supervised learning that includes both classification and numerical regression, which requires a human to label the input data first. Classification is used to determine what category something belongs in, after seeing a number of examples of things from several categories. Regression is the attempt to produce a function that describes the relationship between inputs and outputs and predicts how the outputs should change as the inputs change. Both classifiers and regression learners can be viewed as “function approximators” trying to learn an unknown (possibly implicit) function; for example, a status classifier can be viewed as learning a function that maps signals from a living room, for example, to one or more categories such as normal, danger, or other status.
In an example, the artificial intelligence module can provide for machine to interpret data (from any one of the data sources) in a manner that is similar to the way humans use their senses to relate to the world around them. In an example, the artificial intelligence module takes in and responds to their environment via each of the sensing modules, such as radio frequency module, audio module, inertial motion module, and others. In an example, the artificial intelligence module can also include computer vision, machine hearing, and machine touch. Other examples of artificial intelligence techniques can include, among others, natural language processing, deep machine learning, variations thereof, and combinations, and the like. Of course, there can be other variations, modifications, and alternatives. See, for example, Wikipedia.com, artificial intelligence.
Initially, a number of activities that are desired to be detected are defined, step 2210. In various embodiments, the activities may be relatively simple or complex human user actions, such as: standing up, walking, falling, sitting, running, dancing, rolling, writhing, or the like. Location of these activities may be associated with locations, such as: a user walking in the kitchen, hall, bedroom, etc.; a user sitting in the living room, bathroom, foyer, etc.; a user falling in the shower, kitchen, bedroom, etc. Additionally, in some embodiments, the activities may be actions other than by a human. Some examples of this may include: a refrigerator door opening or closing, a door opening or closing, curtains moving, a rotating fan, various pet movements (e.g. walking, stretching, playing), and the like.
In still other embodiments, activities defined for detection may be non-movement actions that are sensed by other sensors, such as a microphone, a light sensor, pressure, or the like. Some examples of this may include: ringing of an external fire or carbon monoxide detector, phone ring, toilet flushing, waste disposal sound, ice maker/dispenser noise, furnace noise, air conditioner noise, gun shots, specific words spoken by a user (e.g. help, moaning noises, screams, crying, laughter, coughing, sneezing), and the like.
In various embodiments, activities that are defined for detection may also involve combinations of the above activities. For example, a person can be defined to have used a toilet if a human is detected standing in front of the toilet, and a toilet flushing noise is detected; a person can be defined to be cooking if the human is detected in the kitchen, and a microwave hum is detected; a person can be defined to be sleeping if the human is detected in the bedroom, and snoring is detected; and the like. In light of the present disclosure, one of ordinary skill in the art will can imagine that many other types of activities can be defined for detection and are within the scope of embodiments of the present invention.
Next, embodiments of the present invention are typically brought on-site to a test location, and the activities that are to be detected are performed, step 2215. In various embodiments, a variety of test locations may be used, such as different houses, apartments, or the like, and in each location embodiments of the above hardware may be located at different positions therein, step 2220. In some examples, embodiments may be placed in living rooms on fireplace mantles, side tables, consoles, or the like; in bed rooms on dressers or night stands; in kitchens on counters or tables; or the like.
While positioned in each of these locations, the desired activities described above may be performed. For example, while embodiments are positioned on the kitchen table, the various activities discussed above may be performed in a variety of positions within the location, such as a person standing, falling, walking, screaming for help, or the like in the bathroom, the kitchen, the living room, the bedroom, the hallway, or the like.
With each activity performed, step 2225, a variety of data is captured, step 2200. Activity data is typically captured by embodiments of the present invention, such as radar sensor data, audible sound data, and the like, as described above. Typically, such activity data is time stamped. In various embodiments, a video camera (e.g. DV cam, a phone, or the like) is also used to visually record the activity at the same time as when the activity is sensed. This video data is also time stamped. In other embodiments, as an alternative to the video data, data from a wearable sensor (e.g. Fitbit, Apple, Xsens, or the like may be incorporated). Next, the hub device may be located at a different position within the location, and the physical location may also be changed, and the process above may be repeated, step 2235.
Next, in various embodiments, the video data is associated or labelled with the activity performed, step 2230. In some embodiments, this association does not need to be done on-site, but may be done remotely. Typically, this step is performed by a user monitoring the video data, and determining starting and ending times for actions viewed on the video data. For example, t1-t3, t6-t9, and t12-t15 are labelled with “flushing a toilet”; t20-t27, t38-t40, and t45-t55 are labelled with “washing hands”; and the like.
As illustrated in
Next, to help determine a model for activity data, various artificial intelligence/machine learning/deep learning tools are used to determine a model for each activity, step 2250. As can be determined, by recording activity data in a variety of locations and in a variety of positions therein, the activity data captured and labelled with a specific data label can vary dramatically. Accordingly, in various embodiments online tools such as Google Tensorflow, Microsoft Azure, Amazon AWS, Caffe2, Theano, Keras, Torch, and the like can be used to determine a model for each labelled activity. In various embodiments, the number of inputs, and the number of layers used for the decision-making process may vary depending upon speed, accuracy, and other performance characteristics that are desired. The size of the each of the activity models may vary and can be on the order of megabytes.
In various embodiments, once the models for each activity are determined, they are compiled for the target graphics processor (GPU module,
In operation, an end user performs a variety of activities in their home, for example, step 2265, and a variety of incoming data is captured using the hub device, step 2270. As discussed above, the sensed data may include rf data that is captured with the rf sensors; audio data that is captured with the microphone array; positional data captured by an accelerometer; and the like. As discussed above, the incoming data may first be preprocessed by one or more processors, step 2275 e.g. selecting the strongest audio input microphone data, filtering of certain rf data, e.g. multiple source tracking, and the like, using the hardware discussed above. In various embodiments, the processed data is then passed to the GPU module, step 2280.
In response to the processed data, the GPU attempts to identify a data label corresponding to the processed data, step 2285. For example, in response to rf data indicating that the user is standing, the user is in the bathroom, and audio of running water, GPU module may identify an activity associated with the data label “user washing up”; in response to rf data indicating that the user is rolling, the user is in the kitchen, and audio of the user screaming, GPU module may identify an activity associated with the data label “user emergency”; and the like. In various embodiments, the user activity may be identified by a particular flag, or by a value in a register, or the like.
Next, in response to the flag, the MPU or other processor illustrated in
In addition, hub device 800 may provide an indication of the user activity to a remote location (e.g. remote server), step 2297, for handling the activity automatically by the server, or for manual intervention, step 2299. In light of the present disclosure, one of ordinary skill in the art will can imagine that many other types of actions can be performed in response to detection of activities by hub device 800.
In
Next, a pre-processing block 2315 is shown illustrating additional functional detail of a preprocessing process referred to in step 2275, in
In various embodiments, signal processor 2320 performs preprocessing actions on the incoming radar data. In some examples, a DSP, or the like (e.g.
In some embodiments, background estimator 2325 takes the processed incoming radar data and estimates what radar data is associated with static or background signals. In various embodiments, the DSP can perform this function by monitoring the processed incoming radar data over a long period of time, e.g. at night, and then average the data. A result of this step is to attempt to determine what received radar signals are a result of specific placement of a hub device within a room and the specific geographic layout of a room.
In various embodiments, the background data is then subtracted from the processed incoming radar data to obtain radar data associated with moving objects with the room. In the present example, the DSP or MPU (
In some embodiments, the moving object radar data may be captured from multiple objects within a location at the same time. For example, the moving object radar data may capture a fan moving, a dog walking around, a first human walking around, a second human sitting, and the like. Accordingly, the moving object radar data (obtained after subtracting out the background data) may capture multiple moving objects at the same time. In various embodiments, in order to reduce the amount of computational training an AI system will perform to determine activity models; to reduce the memory requirements to store large compiled modeled activities; to reduce the computational work load of a GPU (
More particularly, various embodiments utilize object tracking algorithms 2330 to first identify and track the presence of moving objects from the moving object radar data. In various embodiments, the DSP or MPU (
Next, in various embodiments, based upon the moving object radar data and the object tracking data, the moving object radar data may be extracted into separate object radar data, 2335. In various embodiments, the DSP or MPU (
In various embodiments, the moving object radar data for each object is passed to a pattern recognition engine 2340. As described above in conjunction with
In an example, the present invention provides a system for capturing information from a spatial region to monitor human activities. In an example, the system has a housing, the housing having a maximum length of six to twenty four inches and width of no longer than six inches, but can be other dimensions. In an example, the housing has sufficient structural strength to stand upright and protect an interior region within the housing, but can include variations. In an example, the housing has a height characterizing the housing from a bottom region to a top region and a plurality of levels within the housing numbered from 1 to N, each of the levels configured with one or more modules.
In an example, the system has an audio module comprising a substrate member and a plurality of peripheral microphone devices spatially disposed along a peripheral region of the substrate member. In an example, each of the peripheral microphone devices has an analog output. In an example, the module has a center microphone device spatially disposed within a center region of the substrate member. In an example, the center microphone device has an analog output. In an example, the module has an analog to digital converter coupled to each of the analog outputs. The module has a spatial configuration comprising a circularly shaped region for the peripheral region to provide a 360 degrees field of view for the plurality of peripheral microphone devices. A bus device is coupled to each of the analog to digital converters. In an example, the bus device communicates with each of the plurality of peripheral microphone devices and the center microphone device. The module is coupled to a signal processor coupled to the bus device. The module is coupled to a processor device coupled to the signal processing device and is configured to process an audio information comprising an audio event from the plurality of microphone devices using the signal processors without transferring the audio information to the processing device to achieve a faster selection process of at least one milliseconds to select one of the microphone devices that has a strongest audio signal, and then transfers the audio information from the selected microphone devices. The system also has a cellular network module comprising an interface, which is coupled to the processing device. The system has a user interface configured on an exterior portion of the housing, and coupled to the processor. The user interface allows for a user to initiate and make external calls via the cellular network when desirable or also receive external calls from the network.
In an example, the system has other elements. That is, a speaker device is coupled to the processor device; and an audio driver device is coupled to drive the speaker device. In an example, an LED array is coupled to the processor device. In an example, a plurality of MEMS devices are coupled to the processor device. In an example, a gas sensor device is coupled to the processor device. In an example, a pressure sensor device is coupled to the processor device. In an example, the user interface can be a general purposes input and output device.
In an example, the system has an inertial measurement module comprising an LED array, an accelerometer device, a gas sensor device, and a pressure sensor device configured to detect a pressure within an environment of the housing. In an example, the inertial measurement module comprising a gas sensor to detect a presence of carbon dioxide and coupled to the processor device configured to send out an alert based upon a level of carbon dioxide. In an example, the system has a plurality of LED devices configured spatially around a periphery of the substrate member to allow for illumination of electromagnetic radiation. In an example, the inertial measurement module comprising a i2C bus coupled to a plurality of LED devices, a gyroscope device, an accelerometer device, a compass device, a pressure device, and a gas sensor, the i2C bus coupled to the processing device. In an example, the processing unit comprises an ARM processing unit coupled to a digital signal processor and an image processing unit.
Optionally, the system has a network module comprising an interface, which is coupled to the processing device. In an example, the system has a speaker device coupled to the processor device, and an audio driver device coupled to the speaker device, the processer device being configured with the network module to communicate audio information to output acoustic energy from the speaker device. The system has a user interface configured on an exterior portion of the housing, and coupled to the processor.
In an example, the present invention provides a method of capturing information from a spatial region to monitor human activities. In an example, the method uses an apparatus comprising a housing within a spatial region of a living quarter, which is occupied by a human user or users. In an example, the housing has sufficient structural strength to stand upright and protect an interior region within the housing, the housing having a plurality of levels within the housing numbered from 1 to N. Each of the levels configured with one or more modules, which can include any of the ones described herein and others.
In an example, the housing has an audio module comprising: a substrate member; a plurality of peripheral microphone devices spatially disposed along a peripheral region of the substrate member, each of the peripheral microphone devices having an analog output; a spatial configuration using an edge region for the peripheral region to provide a 360 degrees field of view from the plurality of peripheral microphone devices; a bus device coupled to each of the analog to digital converters, the bus device communicating with each of the plurality of peripheral microphone devices; a signal processor coupled to the bus device; and a micro processor device coupled to the signal processing device.
In an example, the method includes sensing a plurality of audio signals comprising an audio event from each of the plurality of microphone devices. Each of the plurality of microphone device can be receiving an audio signal of a different signal strength based upon a spatial location of each of the microphone devices. The method includes converting each of the audio signals from each of the microphone devices into a plurality of digital signals in a first format using an analog to digital converter. In an example, the method includes processing the digital signals in the first format to a second format, which can be compressed or other form to be transported via an interface. The method includes transferring the digital signals in the second format using a dedicated interface device from each of the plurality of microphone devices into a receive interface device coupled to the signal processing device without transferring the digital signals in the second format to the micro processing device. The method processes information associated with the digital signals using the signal processing device to select one of the microphone devices that has a strongest audio signal as compared to any of the other microphone devices; and transfers information associated with the digital signals from the selected microphone device to an outgoing interface device. In a preferred example, the method includes processing the digital signals from the selected microphone device using an artificial intelligence process to identify the event.
In an example, the present invention provides a system for capturing information from a spatial region to monitor human activities. In an example, the system has a housing. In an example, the housing has sufficient structural strength to stand upright and protect an interior region within the housing. In an example, the housing has a height characterizing the housing from a bottom region to a top region. The housing has a plurality of levels within the housing numbered from 1 to N, each of the levels configured with one or more modules. In an example, the system has a processor device comprising a micro-processing unit, a digital signal processing unit, and an image processing unit. In an example, the system has an audio module comprising a plurality of microphone devices spatially disposed along a region of a substrate member. In an example, each of the microphone devices has an analog output configured to the digital signal processing unit with a dedicated bus. In an example, the system has a radio frequency module comprising a plurality of transmitting antenna and a plurality of receiving antenna and comprising a serial bus coupled to the micro-processing unit. In an example, the radio frequency module is configured to receive back scattered electromagnetic radiation signals from a spatial region; and then transmitting information associated with the back scattered electromagnetic radiation signals to the processor device. In an example, the system has a cellular network module comprising a cellular interface configured to the micro-processing unit with a cellular connection. A user interface is configured on an exterior portion of the housing, and is configured with the processor device to initial a call to an external network using the cellular network module. In an example, the system has an artificial intelligence module coupled to the processor device using a personal computer bus and configured with at least the audio module and radio frequency module to process information from at least one of the audio module or the radio frequency module to process the information to classify the information into one or more classifications; and providing a feedback based upon the classification.
In an example, the system can have other features. In an example, a speaker device is coupled to the processor device; and an audio driver device is coupled to drive the speaker device from a user at a remote location communicating through the cellular network module. In an example, the system has a power management system comprising a power up and down sequence controller, a power good indication, a thermal monitor and protection, a short circuit protection, and a processing health monitor, the power management system coupled to the processing device and coupled to audio module and the radio frequency module. In an example, the system has an Ethernet connection coupled to the processing device. In an example, the system has a plurality of memory resources coupled to the processing device, the memory resources comprising a flash memory array, a dynamic random access memory array, and an EEPROM memory array. In an example, the system has a WiFi sensor module coupled to the processing device using a personal computer interface. In an example, the micro-processing unit comprises an ARM processing unit coupled to the digital signal processing unit and the image processing unit and are configured on a single integrated circuit chip comprising silicon bearing material. Of course, there are other variations, modifications, and alternatives.
In an example, the present invention provides a method of capturing information from a spatial region to monitor human activities. In an example, the method includes using an apparatus comprising a housing within a spatial region of a living quarter. In an example, the method includes sensing a plurality of audio signals comprising an audio event from each of the plurality of microphone devices. In an example, each of the plurality of microphone device receives an audio signal of a different signal strength based upon a spatial location of each of the microphone devices. The method includes converting each of the audio signals from each of the microphone devices into a plurality of digital signals in a first format and processing the digital signals in the first format to a second format. The method includes transferring the digital signals in the second format using a dedicated interface device from each of the plurality of microphone devices into a receive interface device coupled to the signal processing device without transferring the digital signals in the second format to the micro processing device. The method includes processing information associated with the digital signals using the signal processing device to select one of the microphone devices that has a strongest audio signal as compared to any of the other microphone devices and transferring information associated with the digital signals from the selected microphone device to an outgoing interface device. The method includes processing the information using an artificial intelligence module to classify the information into one of a plurality of classifications. Additionally, the method includes processing the digital signals from the selected microphone device and using the artificial intelligence module to identify the event from one of the plurality of classifications.
In an example, the technique transfers learned information and activity information to third parties. The technique teaches itself to learn high level behavior that are indicative of a persons welfare using artificial intelligence techniques. In an example, the present technique will then generate summary of such activities and send it out to the human's loved ones, caretaker or even emergency response team depending on the urgency of the situation. For example for regular days, the technique can simply send short summary like “your mom had a routine activity today”, or “She was much less active today.” In an example, where the human has a care taker visiting few times a week, the technique can send a notification to them, “It seems she struggles more on yesterday”, so that the care taker can pay a visit to make sure everything is fine. Alternatively, the technique can be more acute events like fall, shortness of breathing, or others, that needs quick attention. In these scenarios, the technique can notify medical response team to provide immediate help. Of course, there can be other variations, modifications, and alternatives.
In an example, the present technique can categorize a human target with the listed ADLs, among others. Examples of ADLs including among others, bathing, brushing teeth, dressing, using toilet, eating and drinking, and sleeping. Other ADLs include preparing meals, preparing drinks, resting, housekeeping, using a telephone, taking medicine, and others. Ambulatory activities including among others walking, doing exercise (e.g., running, cycling), transitional activities (e.g., sit-to-stand, sit-to-lie, stand-to-sit, lie-to-sit in and out of bed or chair), and stationary activities (e.g., sits in sofa, stand for a while, lie in bed or sofa). Of course, there can be other variations, modifications, and alternatives.
In an alternative example, the present technique can determine activities of a human target with any one of the activities listed. The listed activities, including among others, and combinations of going out, preparing breakfast, having breakfast, preparing lunch, having lunch, preparing dinner, having dinner, washing dishes, having snack, sleeping, watching TV, studying, having a shower, toileting, having a nap, using the Internet, reading a book, shaving, brushing teeth, telephone, listening to music, doing house cleaning, having a conversation, entertain guest, among others.
In an example, the present technique can also identify a rare event. In an example, the technique identifies when a senior human falls inside a home with no one around. In an example, the technique is robust, without any false negatives. In an example, the technique uses looking at sequence of events that are before to the potential fall and after a potential fall. In an example, the technique combines the contextual information to robustly determine if a fall has occurred. Of course, there can be other variations, modifications, and alternatives.
In an example, the technique also detects and measures vital signs of each human target by continuous, non-intrusive method. In an example, the vital signs of interest include a heart rate and a respiratory rate, which can provide valuable information about the human's wellness. Additionally, the heart rate and respiratory rate can also be used to identify a particular person, if more than two target humans living in a home. Of course, there can be other variations, modifications, and alternatives.
By understanding the context of how the target human (e.g., elderly) is doing, the technique can also provide valuable feedback directly to the elderly using a voice interface. For example, the technique can sense a mood of the human based on sequence of activities and vital signs of the human and then ask, “Hi do you want me to call your son”. Based upon the feedback from the human, the technique can help connect to a third party (or loved one) if their answer is positive. Of course, there can be other alternatives, variations, and modifications.
Having described various embodiments, examples, and implementations, it should be apparent to those skilled in the relevant art that the foregoing is illustrative only and not limiting, having been presented by way of example only. Many other schemes for distributing functions among the various functional elements of the illustrated embodiment or example are possible. The functions of any element may be carried out in various ways in alternative embodiments or examples.
In various embodiments, any number of actions may be taken by the hub device depending upon the type of activity that is detected. In some examples, if a medical emergency is detected, some actions may include automatically communicating with a caregiver or an emergency service (police, fire, security company) via text, telephone, VoIP, or the like, automatically controlling appliances such as turning on a lights, unlocking doors, turning off appliances, and the like; if a natural disaster is detected (e.g. earthquakes, fires, floods) some actions may include automatically turning off gas, water, and/or electrical service, automatically generating a tweet or other social media post, and the like; if a security alert is detected, some actions may also include communicating with a third party, turning on lights, locking interior doors, and the like. In light of the present patent disclosure, one of ordinary skill in the art will recognize many other types of actions and combinations of actions may be performed by the hub device, or by a central server after being notified by the hub device, are within the scope of embodiments of the present invention.
In an example, the present invention provides a method for identifying a status of a user and providing feedback to the user based upon the status of the user. In an example, the method includes generating an electromagnetic radio frequency (RF) signal from an RF generating device and emitting the electromagnetic RF signal into a spatial region of a user. In an example, the method includes detecting a back scattered RF signal, using an RF receiving antenna, from the user and converting the back scattered RF signal from an RF analog signal into an RF digital signal. In an example, the method includes transferring information related to the RF digital signal, using a serial bus, to a micro-processing device coupled to an artificial intelligence processor device comprising a natural language processor and an RF vision processor, which can include one or more combinations of a signal processor and an artificial intelligence processor. In an example, the method includes processing information associated with the RF digital signal using RF vision processor included in the artificial intelligence processor. In an example, the method includes identifying a state of the user based upon the processed information from the RF vision processor and determining a feedback to the user based upon the state of the user. The method includes outputting the feedback using an audio signal from a speaker coupled to a driver coupled to the micro-processing device to the user and sensing, using a microphone device, an audio signal from the user. In an example, the method includes processing the audio signal using the natural language processer to identify a meaning of the audio signal and sensing, concurrently with the audio signal, the back scattered RF signal from the user.
In an example, the RF signal can be selected from a UWB signal, a FMCW signal, or combination of the UWB signal and the FMCW signal. In an example, the RF signal is transmitted by one of a plurality of RF modules configured in a triangular arrangement capable of transmitting the RF signal within a 360 degree range from a center region of the triangular arrangement. In an example, the RF vision processor comprises a plurality of nodes, each of which is tuned with a weighing factor. In an example, the sensing the audio signal and sensing the backscattered RF signal occurs without a visual image processing device, comprising a camera device. In an example, the state of the user is selected from a normal state, a sleeping state, a fall state, in home state, out of home state, cooking state, a bathroom state, an exercise state, a gaming state, an ill state, or an active state or any other state.
In an example, the RF generating device is spatially located in a first room and the user is located in a second room such that the first room is physically separated from the second room. That is, the RF signal can traverse through a physical barrier or other like structure. In an example, the method includes generating a response to the identified audio signal from the user. In an example, the state of the user is associated with a time, a date, a spatial location, and an action of the user. In an example, the state of the user is associated with a vital sign of the user. In an example, the state of the user is provided using an action recognition module, the state of the user is associated with an action of the user. In an example, the state of the user is provided using a tracking module, the state of the user is associated with a spatial location of the user.
In an example, the invention provides a system for identifying a status of a user and providing feedback to the user based upon the status of the user. The system has a housing having a plurality of levels, numbered from 1 to N, where N is an integer greater than 2 and an RF module provided in the housing, the RF module comprising an RF generating source configured to generate an electromagnetic radio frequency (RF) signal from an RF generating device. In an example, the RF module is configured using an transmitting antenna to emit the electromagnetic RF signal through a portion of the housing into a spatial region of a user being in a state. In an example, the RF module is configured using an RF receiving antenna to detect a back scattered RF signal from the user in the state. In an example, the RF module is configured using an analog to digital device to convert the back scattered RF signal from an RF analog signal into an RF digital signal. In an example, the RF module is coupled to a serial bus to transfer information related to the RF digital signal, using the serial bus. In an example, the system a compute module provided in the housing. In an example, the compute module comprises a micro-processing device coupled to an artificial intelligence processor device comprising a natural language processor and an RF vision processor. In an example, the micro-processor device is coupled to the serial bus to receive the information related to the RF digital signal.
In an example, the compute module is adapted to: process information associated with the RF digital signal using RF vision processor included in the artificial intelligence processor; identify a state of the user based upon the processed information from the RF vision processor; and determine a feedback to the user based upon the state of the user.
In an example, the system has an audio module comprising a speaker device coupled to the compute module to output the feedback using an audio signal from the speaker coupled to a driver coupled to the micro-processing device to the user, the audio module being configured using a microphone device to sense an audio signal from the user.
In an example, the compute module is adapted to process the audio signal using the natural language processer to identify a meaning of the audio signal; and wherein the audio module and RF module are coupled to sense, concurrently with the audio signal, the back scattered RF signal from the user. In an example, the RF signal can be selected from a UWB signal, a FMCW signal, or combination of the UWB signal and the FMCW signal.
In an example, the RF signal is transmitted by one of a plurality of RF modules configured in a triangular arrangement capable of transmitting the RF signal within a 360 degree range from a center region of the triangular arrangement. In an example, the RF vision processor comprises a plurality of nodes. In an example, each of the nodes is tuned with a weighing factor. In an example, the sensing the audio signal and sensing the backscattered RF signal occurs without a visual image processing device, comprising a camera device. In an example, the state of the user is selected from a normal state, a sleeping state, a fall state, in home state, out of home state, cooking state, a bathroom state, an exercise state, a gaming state, an ill state, or an active state. In an example, the RF generating device is spatially located in a first room and the user is located in a second room such that the first room is physically separated from the second room.
Also, the functions of several elements may, in alternative embodiments or examples, be carried out by fewer, or a single, element. Similarly, in some embodiments, any functional element may perform fewer, or different, operations than those described with respect to the illustrated embodiment or example. Also, functional elements shown as distinct for purposes of illustration may be incorporated within other functional elements in a particular implementation. Also, the sequencing of functions or portions of functions generally may be altered. Certain functional elements, files, data structures, and so one may be described in the illustrated embodiments as located in system memory of a particular or hub. In other embodiments, however, they may be located on, or distributed across, systems or other platforms that are co-located and/or remote from each other. For example, any one or more of data files or data structures described as co-located on and “local” to a server or other computer may be located in a computer system or systems remote from the server. In addition, it will be understood by those skilled in the relevant art that control and data flows between and among functional elements and various data structures may vary in many ways from the control and data flows described above or in documents incorporated by reference herein. More particularly, intermediary functional elements may direct control or data flows, and the functions of various elements may be combined, divided, or otherwise rearranged to allow parallel processing or for other reasons. Also, intermediate data structures of files may be used and various described data structures of files may be combined or otherwise arranged.
In other examples, combinations or sub-combinations of the above disclosed invention can be advantageously made. The block diagrams of the architecture and flow charts are grouped for ease of understanding. However it should be understood that combinations of blocks, additions of new blocks, re-arrangement of blocks, and the like are contemplated in alternative embodiments of the present invention.
The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense. It will, however, be evident that various modifications and changes may be made thereunto without departing from the broader spirit and scope of the invention as set forth in the claims.
The present application is a continuation of U.S. patent application Ser. No. 17/074,053 (Attorney Docket No. 5924.002US2) filed on Oct. 19, 2020, which is a continuation of U.S. patent application Ser. No. 16/279,954 (Attorney Docket No. 5924.002US1) now U.S. Pat. No. 10,810,850, filed on Feb. 19, 2019; the entire contents of each of which is incorporated herein by reference.
| Number | Date | Country | |
|---|---|---|---|
| Parent | 17074053 | Oct 2020 | US |
| Child | 18434165 | US | |
| Parent | 16279954 | Feb 2019 | US |
| Child | 17074053 | US |
