Patent Application
20190070043
Not Applicable
Not Applicable
Not Applicable
Not Applicable
Not Applicable
Systems for sensing wetness are well documented back to the 1950's where Sears and Roebuck sold a “Wee-Alert” Bed Wetting Alarm. This fact limits or completely eliminates aspects of many patents applications and in fact patents issued since that early 1950's date of sale for a wetness sensor with notification. This puts a real burden of proof on any new patents in the field of wetness sensing to demonstrate or show that the claims made are indeed novel and not just an iteration of prior art that would be obvious to one familiar or skilled in the subject material.
There is also a distinct difference between an wetness notification system and a wetness sensing system. A notification system, merely would acquire data from a sensor which is developed and designed to detect wetness and process that wet notification to a caregiver or user. On the other hand, a wetness sensor would be the mechanics and methodology of properly determining a wet event and feeding that notification into a wetness sensing system for processing. These two concepts are distinctly different and must be treated as such. A couple of examples of the complexity of sensing wetness determined during over 20 years of development would demonstrate this. In early trials of a wetness sensing system prior to the filing of the Collette et al [US 2005/0033250] at Kimberly Clark, failures to properly sense wetness in non-woven disposable diapers led to the discovery of the impact of non-woven material in the generation of static in the diaper leading to significant failures in current wet sensing capabilities. That led to the development and patent of algorithms and methods to remove static events as triggers for wetness sensing. Another example of the criticality of sensors in wetness systems was one that led to the filing of this paten application. While implementing a resident monitoring system for nursing homes, assisted living centers and hospice, a higher than expected ratio of false alarms was detected in the hospice application. Analysis of the data showed that as the resident became increasingly ill, the decline in kidney function led to a significant increase in mineral content excreted by the kidneys. This resulted in the sensor as designed being unable to correctly measure an incontinent event even with designed in level adjustments. These factors, even to one experienced in the art, would not lead to the development and patentability of critical sensor design modifications, without significant investment of time and material resources, which are what patents are intended to protect.
In all prior art, the sensing of wetness, and that includes wetness sensing clear back to the 1950's Sears and Roebuck, “Wee-Alert” alarm system, has been done by threshold levels. A sensor sends a resistance, capacitance or other measure through an analog comparator which has a set of threshold(s) which then trigger a wet notification. Most often, these threshold levels are changed by manual input and require changes to board configurations or software that changes the comparator to a different resistor via a mux. Digital resistors became available around 2004. Even through they have been available, there have been no wetness sensors to date that make use of that technology in their systems. This patent is that the sensor is not only adjustable via the digital resistor but that adjustment is achieved automatically by the sensor itself.
Not applicable (See the Detailed Description of the Invention)
The current versions of wetness sensors as described in U.S. Pat. No. 8,421,636 B2 issued Apr. 16, 2013 do a very good job of detecting, without false alarms, the incontinent product users initial wet event in a dry incontinent product. The ability to detect subsequent wet events and diaper saturation is a more challenging problem due to the wide range of user urine salinity which is the main factor in the urine resistance/conductance level.
Following the initial wetness event, the ability to dynamically report the exact resistance value of the garment after the wet event and then set wetness sensing unit's new threshold resistance level based on this user's urine resistance enables a new class of wetness sensing capabilities. It will enable the sensor to better track a user's current wetness and detect subsequent wet events temporally correlated to the user's current hydration and health. Providing this data back to the cloud also enables long term tracking of a user's urine salinity which could potentially enable the detection of a change in the user's health and potentially predict and or catch an issue well before it is evident by other means.
The temporal correlation of a user's urine resistance should be very close from wet event to wet event but rate of change resistance of the users urine will be measurable based on temperature, i.e. lower resistance immediately after incontinent event and it will reduce as urine cools within the incontinent product. Utilizing the dynamic tracking function of the new wetness sensor should enable the ability to detect on new wet events and also track overall diaper saturation based on time since first event, subsequent events and delta change in resistance.
The addition of a digital potentiometer to the wetness sensor enables the ability to change comparator thresholds based on in-system and real time measurable parameters but still enable the ultra low power operation of comparator based sensing.
With the dynamic configuration of the new wetness sensor married with a system level enabled tracking and reporting system the wetness detecting system now has the ability to more intelligently report initial and subsequent incontinent events, over all diaper saturation and potentially the incontinent users heath (as it pertains to the urinary tract) Adding additional measurement systems, such as temperature, capacitance, ph could help in the detection determination of a secondary wetness events but will ultimately drive up the complexity and cost of the diaper and sensor and making it less affordable. These additional sensing elements will be considered more in the future.
The following section will cover the implementation of this new technology into the System.
Wet Sensor Dynamic Algorithm Description
The algorithm updates to support the new dynamic capability of the wet sensor includes updates to the Wet Sense Unit firmware and the System level monitor (local webpage).
A high level overview of the new algorithm flow for the Wet Sense Unit:
Detailed updates to enable dynamic sense capability to the wet sense algorithm are captured below.
Overview
In prior version of the Wetness sensor the threshold values of the sensor were set by discrete resistance components. This enabled very accurate initial wetness detection but for subsequent incontinent events the fixed resistance components did not allow for an adaptable and dynamic reconfiguration. The new sensor utilizes a digital potentiometer to facilitate this new adaptable and dynamic reconfiguration.
The analog devices AD5165 is a digital potentiometer and has been integrated into the design and will enable this new function but any digital potentiometer would suffice.
The digital potentiometer is 100K Ohm end to end, between port A and B. The sensing will take place off of port W which we can configure between 100K to 0 Ohms. There are 256 resistance steps between 100K to 0 Ohm, this is represented by a 8 bit digital value, resulting in a resolution of 390 Ohms per step.
This enables the sensor to determine urine resistances of less than or equal to 50K Ohm at 390 Ohm of resolution per step (Current=7 mA per step, Conductance 2.5 mS per step) This results in a dynamic resistance command word format of: bXXXX_XXXX, where the most significant bit X represents a configurable value of 1 or 0.
This results in a control dynamic control range of:
B1111_1111 (˜100 kohm)) to b000_0000 (˜0 Ohm)
Initialization:
The initialization of the algorithm does not change between this version and past versions.
There are three types of initialization events: power on reset, magnet/proximity swipe connection check or the periodic system connection check.
Initialization: Power on Reset:
The power on reset initialization is executed when the battery is applied. The digital resistor is set to 50 KOhms initializing the sensing threshold to a dry garment.
Initialization: Magnet/Proximity Swipe Connection Test
During an incontinent product change the caregiver will swipe the Wet Sense unit with a magnet. This will trigger a diaper connection test event. The connection event currently checks to see if the diaper is connected correctly to the garment.
As part of this algorithm update the initial threshold/resistance value will also be dynamically set.
Magnet/Proximity Swipe Connection Test Algorithm flow:
The Message format will be updated to include the current resistance setting/threshold.
Connection Status:
At the system level, if the value is 50K Ohm it signifies the garment is dry. If the resistance value is less than 50K Ohm, it may indicated that the garment is already wet. This information will be used by the system to determine notification type.
Initialization: Periodic Connection Test
The periodic connection test executes every TBD minutes to check that the diaper is still connected correct and to check the wetness of the product and set the sensor back to 50 KOhm if garment was changed but not swiped.
This algorithm is similar to the magnet swipe connection test algorithm except the only check performed against the diaper resistance is if dry diaper threshold of 50 KOhm is valid, if it is not the threshold is set back to the current threshold
Magnet Swipe Connection Test Algorithm flow:
The Message format will be updated to include the current resistance setting/threshold.
Connection Status:
The periodic reporting the incontinent product connection status and current threshold level enables monitoring of the sensing system and incontinent product's state of health.
Monitoring Operation:
After the initialization phase of the sensor unit it will enter monitoring mode. There are two monitoring modes of the wet sensor; edge monitoring and periodic level monitoring.
The reason for having the two modes is power conservation which directly impacts battery life. The updated wet sense units ability to dynamically sense and track subsequent wet events over a very large dynamic range requires the sensor unit to configure the wet sense threshold/resistance to levels low enough to enable sense additional wet events. In older versions of the wet sense unit its dynamic range was very limited and the unit would routinely saturate after the first wet event making it unable to detect additional wet events.
The power required to monitor increases inversely to the threshold setting. As the incontinent product becomes more saturated the threshold/resistance level lowers and the power required to monitor at that new threshold goes up.
To maintain battery life if the threshold/resistance drops below TBD KOhms the unit will switch between the always on, always monitoring, edge monitoring mode to the periodic level monitoring mode.
These two monitoring modes functions and their differences are captured below.
Monitoring: Edge Monitoring
The edge monitoring mode of the wet sensor is the default mode of the sensor, it is also the historical/classical sensing mode and has been utilized since the very first versions of the wet sense model. The implementation of this mode is not changing and only and only an implementation overview will be provided in this document.
Edge monitoring mode is always on and always monitoring enabling the wet sense unit to capture wet events real time, as they are occurring.
High level flow:
The power consumed by the comparator and micro controller while in the sleep mode is very low.
The power utilization increase comes in the form of the resistor divider created by the pull up threshold resistor and the pull down diaper impedance. As the diaper impedance/resistance drops so does the pullup threshold resistance increasing the amount of current that can flow between VDD and Ground.
Monitoring: Periodic Level Monitoring
When both the Diaper Impedance and threshold/resistance (Digital Resistor in figure above) drop below TBD K Ohms the sensor will need to cut off this constant current supply and enter its periodic level monitoring mode.
The periodic level monitoring mode has two phases:
When in this mode the sleep phase and the level monitoring phase are repeated over and over at a periodic interval, hence the name periodic level monitoring.
The configuration flow of the periodic level monitoring mode is:
Wet Event Processing
If either the edge monitoring and periodic level monitoring modes trigger a wet event both will enter the wet event processing function to transmit a wet event message and set the unit back up for additional monitoring.
The Wet Event Processing Algorithm Flow:
Wet Event Message Format:
In the current wet event message the 8 bit wet count field that is currently not used.
The current wet module message format is shown in the table below
To transmit all the information required for the new dynamic sensor but to minimize message length the currently unused “Wet Counter” byte will be repurpose and add an additional byte will be added. The new message format definition will be defined below.
Note, adding a byte to the overall message length is not significant because all zigbee message payloads utilize delimiter “̂” between each field enabling dynamic message length functionality. So repurposing and extending the length of the Wet counter will only impact that part of the system processing and all other field processing will remain unchanged.
Below is the new wet event message format. The Wet Count byte has been replaced with a two byte field labeled “Wet State”
The new message field “Wet State”, reports the current event type and pre and post wet resistance values.
The Wet State field is defined below.
Triggered Wet Resistor Value: Is the digital potentiometer 8 bit register value that maps to a resistor value that the unit triggered on to generate the current Wet Event message.
New Wet Resistor Value: Is the new potentiometer 8 bit register value that the unit measured after the wet event. This value maps to the resistor value measured in the unit that does not result in the comparator to trip but is very close to this point.
Note the “New Resistor Value” Is NOT the new threshold value the sensor unit is set to sense subsequent events. The new Threshold value will be a count/resistance value TBD KOhms lower to create a bit of margin/hysteresis against detecting the next event and not false alarms on an already wet incontinent product.
8
Network NET Description
The Raspberry PI code will be updated to support the message decode of the new wet sense messages but still post the message to the local system website such that the system still works as designed today.
The code will be updated again once the system level web page has been updated to support the new message formats.
Detailed message formats are captured Monitoring and Wet Event sections above.
System Level Wet Sense Description
To support the new dynamic wet sense
The System level algorithm flow:
Configuration
The configuration control of the AD5165 will be captured next.
Configuring the W port requires the development of a serial port driver. Refer to
See
See
See
Not Applicable
