Patent Application
20240417938
The invention is in the field of a mobile installation for street cleaning, also referred to as a cleaning machine for a road or pavement or gutter, for removing undesirable matter from roads or like surfaces, with or without moistening of the surface. These machines are used to remove objects, such as small items, such as leaves, and debris, and dirt from a road or the like, using a broom system for cleaning the road or the like, and typically a storage for removed objects and dirt.
The present invention relates to a cleaning machine. A main function of such a cleaning machine is to remove dirt/debris specifically from the gutter (side of the road, edge of the pavement; curb) and to remove it from the street; dirt/debris may be understood to relate to all smaller items being present and to be up-taken. Collecting and removing the debris is typically done by means of broom(es) which sweep the debris in a direction of an uptake element, and uptake of the debris then takes place, such as by under-pressure, e.g. vacuum.
For proper up-taking of dirt/debris an operator, typically the driver of the cleaning machine, uses a broom control device to move the brooms towards the dirt and debris, and at the same time steering the cleaning machine, paying attention to traffic and pedestrians, and so on. The broom control device may be a joy stick or the like. The position of the entire machine is controlled by the steering controls of the driver, such as a typical steering wheel. It is found that manually and simultaneously controlling all these movements requires a high level of practice, as well as good eye-hand coordination, and concentration. In practice, there are very few operators that are skilled enough to get optimum performance of the system. Lack of accuracy of the above described manual process may have a large impact on quality, efficiency, and speed of the cleaning work performed. Also, from an ergonomics point of view, the above is a challenging topic to address, as it is physically demanding and technically very complex to solve.
Some documents may be referred to. For instance DE 10 2017 117013 A1 recites a semiautonomous following device, having at least one drive apparatus, at least one apparatus for steering the drive apparatus, at least one apparatus for detecting a guidance signal, at least one control unit which activates, as a manipulated variable, the apparatus for steering the at least one drive apparatus, as a function of the guidance signal. According to the invention there is provision that the control unit or a unit which is separate from the control unit activates further apparatuses as a function of the guidance signal, wherein the further apparatuses are selected from the group comprising: direction indicator, vehicle lights, machine modules and radio-operated remote-control systems. The interconnection of the apparatuses increases the safety of the locomotion of the following device in the public space and also the efficiency of the cooperation between the machine and the user. CN 103 161 133 A recites a an intelligent road-cleaning vehicle based on machine vision. The intelligent road-cleaning vehicle comprises a vehicle body provided with a cleaning machine, a first light source, a second light source, an image obtaining device, an image extracting device, an image processing device, a first judging device and an automatic turning device, wherein the image extracting device is used for extracting an image of a targeted area from a road surface image, the image processing device is used for conducting grayscale adjustment and binarization treatment to the image of the targeted area to generate a binary image and covering a two-dimensional array corresponding to the binary image to count the number of elements with values of one in the two-dimensional array, the first judging device is used for judging whether the number of the elements with the values of one is larger than a preset value, and the automatic turning device is used for sending an order of turning right to the vehicle body when a judging result of the first judging device is yes, and sending an order of turning left to the vehicle body when the judging result of the first judging device is no. The intelligent road-cleaning vehicle is based on the machine vision, is capable of turning and changing speed automatically, and has a value of actual popularization. CN 103 696 381 A recites a control method, a control device, a control system and a sweeping vehicle for road edge cleaning. The control method comprises the following steps: obtaining the distance between a sweeping device and a road edge in real time according to the information of the road edge; when the distance between the sweeping device and the road edge exceeds a preset range, sending out a control command so as to adjust the position of the sweeping device, so that the distance between the sweeping device and the road edge is always in the preset range so as to ensure that the sweeping of the road edge is efficiently completed, avoid collision with the road edge, reduce the working difficulty of operators and improve the work efficiency. WO 2020/144011 A1, which is published after the present priority date, recites a method for operating a cleaning vehicle, in particular a road sweeper which can be operated in an automated manner, by means of a control device, wherein a trajectory is calculated and control commands for longitudinal guidance and lateral guidance of the cleaning vehicle when driving along the calculated trajectory are generated, measurement data are received from at least one sensor for determining a cleaning area, and at least one actuator for positioning a cleaning apparatus of the cleaning vehicle for cleaning the cleaning area is controlled in the longitudinal direction and/or transverse direction relative to a direction of travel of the cleaning vehicle on the basis of the received measurement data. A control device, a computer program and a machine-readable storage medium are also disclosed.
Cleaning machines are capable of producing lots of noise. They also consume significant amounts of energy.
Incidentally WO 2020/152526 A1, WO 2021/118341 A1, and CN 112 211 145 A may be mentioned. WO 2020/152526 A1 recites an automated road sweeper comprising means of transport suitable to allow the sweeper to travel on a road surface including debris, a plurality of cleaning members arranged at fixed points located in different portions of the sweeper and suitable for cleaning within predetermined cleaning areas, and selection means operatively connected to the cleaning members, suitable for scanning the road surface and locating the debris so as to select and exclusively operate the cleaning members, the cleaning area of which covers the debris during the movement of the sweeper. WO 2021/118341 A1 recites a cleaning machine for a road or pavement or gutter. These machines are used to remove objects, such as small items, such as leaves, and dirt from a road or the like, using a broom system for cleaning the road or the like, and typically a storage for removed objects and dirt. CN 112 211 145 A recites a semi-automatic road sweeping method and device for a road sweeper. The method comprises the steps that a color image and a depth image of a road surface in front of a vehicle are collected through a vehicle-mounted binocular camera installed on the road sweeper; pictures under various road conditions are collected, and a data set of the road surface cleanliness is made and used for training a convolutional neural network model; a real-time road condition image collected on the road sweeper is sent into the trained convolutional neural network for calculation, and a vector of a road surface cleanliness degree is output; according to the depth image information, obstacles which hinder passing and appear in the passing space in the advancing direction of the road sweeper are detected; the width of the road surface in front of the road sweeper is calculated according to the depth image, and the road sweeper is guided to automatically inspect edges; the energy consumption of the road sweeper is greatly reduced, the automation degree of the road sweeper is improved, and a driver is assisted in conducting road sweeping operation.
The present invention therefore relates to an improved cleaning machine, which is more energy efficient, more silent, and with an equivalent amount energy can clean more road or pavement, which overcome one or more of the above disadvantages, without jeopardizing functionality and advantages.
The present invention relates in a first aspect to a cleaning machine (1) for a road or pavement or gutter comprising a vehicle (2), the vehicle comprising a suction system for up-taking debris and dirt from a suction point, such as an enlarged suction opening, connected to the suction opening a suction tube, at the other end of the suction tube a container for receiving dirt and debris, and a fan for providing suction, such as by under-pressure, and at least one broom system (3), wherein brooms are adapted to sweep debris and dirt to a suction point, such as by rotating, at least one camera (7) providing images, typically a 2D or 3D camera, pattern recognition software adapted to process the images, wherein the pattern recognition software is configured to identify physical characteristics of debris and dirt to be up taken by the suction point, and at least one controller (4) for operating said broom system (3), characterized in that the at least one controller (4) is configured to increase or decrease suction by controlling said fan in view of the physical characteristics of debris and dirt to be up taken by the suction point. The at least one camera is typically positioned such that debris and dirt in front of the present cleaning machine are recognised, such as by placing it at a front side of the present machine. The suction may be varied from 10-100% of a maximum, for suction of lightweight material to suction of relatively heavy material, in particular from 20-90% of a maximum, such as from 25-80% of a maximum, or stepwise from 5% in steps to 100% of said maximum, such as in steps of relatively 1% increment or decrement, respectively. Suction power is typically controlled by increasing or decreasing power of a fan or the like, such as by controlling the rotation speed from typically 1000-4000 rpm, more in particular from 1500-3000 rpm. At low speed the suction is typically some −1 kPa, whereas at maximum speed suction may be in the order of −20 kPa, depending on the suction device. Therewith the present cleaning machine is more energy efficient, more silent, and with an equivalent amount energy can clean more road or pavement. The at least one vertically and horizontally movable broom system is typically provided on a right side of the cleaning machine in countries where vehicles drive at a right side of the road, or in the alternative on the left side. A second broom may typically be provided at the other (left or right) side. So each broom system may comprise one broom, two brooms, or more than two brooms, such as three brooms. Therewith significant improvements are achieved. By implementing a camera, image recognition and connecting it to the described machine control functions, the above described functions are automated during a significant part of the operating/cleaning process. A reliable, automated broom positioning system is found to have a positive effect on quality, efficiency, speed and ergonomic aspects of the work performed. A reliable recognition of the curb is provided, by ignoring the dirt (to be removed) visible in the image, compensating for broom-wear, and adapting for difficult image situations (high contrast, low light, etc.). Added value of the system is found in a lower strain, stress, and fatigue for the operator during significant part of the operation, both physical (ergonomics, repetitive movements, sitting position) and psychological (concentration, fatigue), a better and more constant quality of work, i.e. better and more efficient cleaning, and no declining of quality by loss of concentration or distraction of the operator. Further it is found that past experience of the operator has less or no influence on the quality of the work. In addition, an improved traffic security is provided, as the operator can pay more attention to traffic, pedestrians and other surroundings. Typical components of the present pavement track system are a 3D camera, image recognition software+hardware, including a connection to machine control functions. The image recognition software used may be Open Source software. The software is taught to detect the curb. Electronic control of at least one broom side-shift is provided, as well as electronic control of vehicle steering (including manual override safety). Some main software steps may be acquisition of a 3D image from the camera, a translation of the 3D-image to a 2D image, a determination of a “rising edge” or “falling edge” in the 2D image, representing the curb, identifying at least 3 points on the “rising edge”/“falling edge” and thereby validating a line representing the curb on reliability, such as by calculation a standard deviation, establishing if these at least 3 points are reliable for at least 2 consecutive frames/images and then accepting the representation of the curb, and further, when accepted, these points are sent to a machine control such as via CAN, to initiate the required control actions. It is found that the identification of the falling edge, i.e. an upper edge of a curb, provides more reliable and reproduceable results. The edge is preferably determined ahead of the machine, in the direction of movement, such as 0.5-3 m ahead of the machine, e.g. 1-2 m. As the speed of the cleaning machine is known, the speed of the machine, and the measured distance of the broom from a curb, can be used to move the brush-system towards or from the curb when advancing. Typically positioning of the can be done by moving the broom-system and/or vehicle sideways such that a distance between the broom system and side of the road is controlled. The vehicle control can at least partly by overtaken by an automated system, in view of the present hydraulic system, such as an orbitrol. The automated system can perform relatively simple tasks, such as positioning of the cleaning machine, driving, steering, and somewhat more complex tasks, as recognition of obstacles and human beings, especially in front of the machine. In view of the limited speed of the cleaning machine such an automated system could already be sufficient.
For improved range and added functionality, an option is to use the third broom system. This option provides an increased sweeping width/range and is often used with a dedicated broom for weed removal. Operating this third broom typically has similar requirements as the first and second brooms: it may be too far from the curb and thereby dirt/debris is missed/not picked up, it may be too close to the curb and the wires of the broom are pushed flat, the broom does not move the dirt effectively. Operation of the third broom may be done by a second joystick, in addition to the standard controls. For the operator this implies operating two joysticks simultaneously; operating/movement of this 3rd broom is very frequent. The construction of the third broom is typically similar to the first and second broom system: the position of the third broom relative to the machine, controlled by the so-called “third broom side-shift”, this is a linear movement, perpendicular to the length axis of the machine (driving direction). And the distance between the machine itself and the curb is controlled.
In a second aspect the present invention relates to a pavement track system computer program comprising instructions for operating the cleaning machine (1) according to the invention, the instructions causing the computer (1) to carry out the following steps: loading 3D images, processing the 3D-images with pattern recognition software, wherein the pattern recognition software is configured to identify physical characteristics of debris and dirt to be up taken by the suction point, obtaining said physical characteristics, and increasing or decreasing suction by controlling said fan in view of the physical characteristics of debris and dirt to be up taken by the suction point.
Thereby the present invention provides a solution to one or more of the above-mentioned problems.
Advantages of the present invention are detailed throughout the description.
The present invention relates in a first aspect to a cleaning machine according to claim 1.
In an exemplary embodiment the present cleaning machine comprises a pavement track system (6) comprising a (further) 3D camera (7) providing 3D-images, pattern recognition software adapted to process the 3D-images, wherein the pattern recognition software identifies a side of the road, and wherein the pattern recognition software measures a distance from the vehicle to the side of the road, and wherein the at least one controller is adapted to move the broom-system sideways with respect to the vehicle such that a distance between the broom system and side of the road is controlled. The same camera as for identification of physical characteristics may be used, or at least one additional camera. Likewise the same controller may be used, or at least one additional controller. By using the pavement track system the present cleaning machine can be operated semi- or fully automatically.
In an exemplary embodiment of the present cleaning machine brooms are adapted to move with respect to the cleaning-machine in a horizontal direction and/or vertical direction, preferably move each individually, and preferably move independent in the horizontal and vertical direction. Therewith amongst others broom wear is minimized.
In an exemplary embodiment of the present cleaning machine the at least one controller is adapted to move the broom-system sideways with respect to the vehicle such that a distance between the broom system and side of the road is controlled, in particular wherein the vehicle comprises at least one seat for an operator.
In an exemplary embodiment of the present cleaning machine the physical characteristics of debris and dirt to be up taken are selected from size, volume, chemical species, weight, density, colour, humidity and/or moisture content, and combinations thereof. For instance, wet leaves have a higher density that dry leaves, and therefore suction power can be adapted in view thereof.
In an exemplary embodiment of the present cleaning machine suction is decreased in case of low density objects, or likewise lightweight objects, such as decreased by 1-20 kPa, and wherein the suction is increased in case of high density objects, or likewise higher weight objects, such as in case of sand, pebble, and dirt, such as increased by −1 to −20 kPa. So for instance if mainly leaves, plastic bags, or the like are detected, then suction is decreased, whereas if pebbles or the like are detected, the suction is increased. The increase or decrease is typically maintained for 1-10 seconds, such as 2-5 seconds, until the objects are sucked in, and then suction is returned to “normal” operation mode. A normal operation mode is typically at 7-80% of a maximum suction power.
In an exemplary embodiment of the present cleaning machine the fan is adapted to provide a suction power pressure of −70 kPa to −1 kPa, in particular −30 kPa to −2 kPa, more in particular −20 kPa-−5 kPa, such as −6-−3 kPa.
In an exemplary embodiment of the present cleaning machine the fan is adapted to provide a suction such that debris or dirt is uptaken through the suction tube with a speed of 1-100 m/sec, in particular 5-40 m/sec.
In an exemplary embodiment of the present cleaning machine the fan is adapted to reduce noise by 3-20 dB, in particular 5-8 dB.
In an exemplary embodiment of the present cleaning machine the controller is adapted to move the vehicle sideways such that a distance between the broom system and side of the road is controlled.
In an exemplary embodiment of the present cleaning machine the controller adapts the distance in view of broom wear.
In an exemplary embodiment the present cleaning machine further comprises a second broom system, and optionally a third broom system.
In an exemplary embodiment of the present cleaning machine the software adapts for contrast and/or wherein the software adapts for shadow effects.
In an exemplary embodiment of the present cleaning machine the software is trained for recognizing the pavement, such as for identifying a rising edge thereof (curb).
In an exemplary embodiment the present cleaning machine is adapted to be steered by an operator or driver.
In an exemplary embodiment of the present cleaning machine the software is adapted to validate a position of the rising edge, such as by validating if at least three identified points of the rising edge have a linear relationship (are on one line), and/or wherein the rising edged is validated on at least two consecutive 3D-images, and combinations thereof.
In an exemplary embodiment of the present cleaning machine the control is electronic, and/or wherein a CAN-bus is used for control.
In an exemplary embodiment of the present cleaning machine the controller is adapted to be manually over-ride.
In an exemplary embodiment of the present cleaning machine the controller is adapted to maintain the distance between the broom system and side of the road at a constant value within a predetermined limit.
In an exemplary embodiment of the present cleaning machine an orientation of the 3D-camera is adaptable, such as adaptable to be oriented substantially towards a side of the road, and/or adaptable to be in a height position providing images of the side of the road.
In an exemplary embodiment of the present cleaning machine the cleaning machine comprises at least one ventilation opening, and an outlet, preferably comprising a filter for removing small particles.
In an exemplary embodiment of the present cleaning machine comprises a display (5).
In an exemplary embodiment the present pavement track system computer program further comprises instructions for compensating for broom-wear, and/or adapting for difficult image situations, such as high contrast, and low light.
The invention is further detailed by the accompanying figures and examples, which are exemplary and explanatory of nature and are not limiting the scope of the invention. To the person skilled in the art it may be clear that many variants, being obvious or not, may be conceivable falling within the scope of protection, defined by the present claims.
In the figures:
The figures are further detailed in the description and examples below.
The present cleaning machine takes over a significant part of the manual control of several functions of the sweeper, during sweeping operation. This reliefs the driver of a difficult, tiring task.
In an example the present modification is designed and integrated in the sweeper, leaving all existing/present functions of e.g. the Ravo type 5 machine otherwise unchanged. Controlling these functions can still be performed manually, but also an automatic mode is available.
The present sweeper, or streetsweeper, can best be described as a large domestic vacuum cleaner. All typical components of a vacuum cleaner are typically there: an engine driving a (centrifugal) fan that creates vacuum, thus creating air movement in a suction nozzle that picks up dirt/debris. Through a (suction) tube the airflow transports the dirt/debris into the dirt/debris container (See
The above describes functionality of the present sweeper, relating to proven Ravo 5 technology.
The integration of the present invention into the Ravo 5 machine may be done mainly electrically, such as through software, or by adding an electrical function to an existing (hydraulic) component. An example of the latter is an additional, electrically controlled steering function. This allows the electronic machine control to influence the mechanical/hydraulic steering unit, thus creating automated steering.
Addition of the 3D camera to the machine, creating a 3D image to determine the distance between the machine and the curb, by image recognition, and bring this information with respect to the distance to the control-computer of the machine, allows the introduction of an automatic sweep function, relieving the driver of a tiring task, creating better efficiency, and more safety.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2030142 | Dec 2021 | NL | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/NL2022/050712 | 12/9/2022 | WO |
