DE102020207782A1 - Method for determining an emergency trajectory - Google Patents

Abstract

The invention describes a method for determining an emergency trajectory for an automated robot and comprising the following steps: • receiving (102) environment data; Determining (104) a deceleration value for the detected object and / or the area that can be driven by the robot; determining (105) an emergency trajectory through the determined object and / or along the area that can be driven by the robot based on the determined deceleration value.

Description

The invention relates to a method for determining an emergency trajectory for an automated robot, a device set up for executing the method, a computer program for executing the method, and a machine-readable storage medium on which this computer program is stored.

State of the art

Methods for determining emergency trajectories are known from the prior art. For example, the DE 10 2016 211 587 A1 a method of controlling a vehicle. In this case, a collision signal is read in, which represents an impending collision of the vehicle with at least one collision object. An emergency trajectory is then selected using the collision signal.

Disclosure of the invention

• • Empfangen von Umfelddaten;
• • Erfassen wenigstens eines Objektes und/oder eines mittels des Roboters befahrbaren Bereichs basierend auf den Umfelddaten;
• • Ermitteln eines Verzögerungswertes für das erfasste Objekt und/oder den mittels des Roboters befahrbaren Bereich;
• • Bestimmen einer Nottrajektorie durch das ermittelte Objekt und/oder entlang des mittels des Roboters befahrbaren Bereichs basierend auf dem ermittelten Verzögerungswert.

The invention describes a method for determining an emergency trajectory for an automated robot and comprises the following steps:

• • Receiving environmental data;
• • Detection of at least one object and / or an area that can be traveled by the robot based on the surroundings data;
• • Determination of a deceleration value for the detected object and / or the area that can be traveled by the robot;
• • Determination of an emergency trajectory through the determined object and / or along the area that can be driven by the robot based on the determined deceleration value.

The environment data can be, for example, data of the robot environment recorded by means of a sensor. The sensors can be, for example, robots' own sensors, such as video, radar, lidar and / or ultrasonic sensors. It can also be acoustic data from microphones. In particular, it can be data from a camera which has a high sensitivity in the infrared range, in particular in the far infrared range. The sensors can also be attached to other robots and / or to infrastructure facilities.

In addition, the environment data can also be data from other robots and / or an external server that have already been evaluated. For example, several data received from this server can already have been aggregated and evaluated on a server.

The object can be any object in the detection area. Objects of interest are preferred which the vehicle could still drive at the time of detection and which consequently lie in a possible emergency trajectory of the vehicle. The recorded area is preferably also an area that is located along a possible emergency trajectory.

The object can be, for example, bushes, hedges, trees, street signs, other vehicles, in particular parked vehicles and / or vehicles without occupants, walls, houses, street lamps, crash barriers or the like.

The areas can in particular be areas away from the intended traffic routes. Possible areas are green areas, arable land, ditches, bodies of water of all kinds, sandy areas, sidewalks and the like.

Various methods can be used to determine a delay value for a detected object and / or a detected area. For example, the size, dimensions and weight of objects can be estimated and probable delays can be determined on the basis of these values and the law of conservation of momentum. The rigidity of objects, for example in combination with ground anchors, for example relevant for plants and infrastructure facilities such as delineator posts, street lamps, traffic lights and signs, can be used as the basis for calculating a delay of the robot. The rigidity plays a decisive role in determining the acceleration of the robot, especially in the case of very heavy objects. Instead of estimating the data of objects, measurement data, in particular images of the objects, can also be compared with a database and, as a result, more detailed information about the objects can be determined. In particular, algorithms that have been trained using machine learning processes are ideal for classifying objects and areas.

For the determination of a deceleration value of an area, for example, the coefficient of friction of the areas can be estimated or determined and serve as a basis for calculation. This can be done inside the vehicle or external to the vehicle, for example by comparing captured images with corresponding databases. For more complex substrates, more complex models, simulations or algorithms trained using machine training methods can be used.

The determination of the emergency trajectory is based on the determined deceleration value. Ever Depending on the current driving situation, different delays may be required. If people and / or other living beings are transported by the robot, predefined acceleration limit values should not be exceeded, so that serious injuries or even the death of people or other living beings can be prevented. In most emergency situations, however, the aim is to get a robot into a safe position as quickly as possible, which in many cases can be guaranteed by stopping the robot. Thus, when determining the emergency trajectory, a primary goal can be, for example, to bring the robot to a standstill as quickly as possible, with predefined acceleration threshold values not having to be exceeded.

However, if there is a risk of injuring other living beings that are or could be along the trajectory, the threshold values can also be adjusted so that the overall damage to all road users / living beings is minimized.

If the robot does not transport living beings or hazardous substances, such as explosives, acids, fuel or the like, there is no danger of harm to people or living beings when the robot is decelerated as quickly as possible. There are situations in which it can be advantageous to decelerate the robot to a standstill as quickly as possible without adhering to predefined threshold values. This is possible in cases where the robot is not transporting living beings or hazardous substances. Then, for example, a robot designed as an automated vehicle could be steered to a tree in an emergency situation so as not to endanger people and / or other living beings in its vicinity.

The automated robot can be an at least partially automated robot, which can be designed as an industrial robot, lawn mower, watercraft, rail vehicle, motor vehicle, bus, shuttle, driverless vehicle, aircraft, drone and / or helicopter. In particular, the robot is an assisted, partially, highly or fully automated vehicle. The vehicle is preferably controlled in a highly or fully automated manner and can also be operated completely without the intervention of a human driver.

The method described here offers the advantage that the safety of vehicle occupants and other people, in particular road users, is increased. By continuously determining an emergency trajectory while an automated robot is driving, a predefined safety measure can be maintained during operation, which can ultimately lead to a release of a correspondingly automated vehicle. By actually following a determined emergency trajectory, injuries, damage and dangerous situations for living beings around the robot can then be reduced or avoided in an emergency.

One of the aims of the method is to provide a system that determines the trajectory with the likely lowest crash severity and makes it available when a “classic” trajectory on a traffic route provided for the robot and / or a trajectory without a collision with other objects is carried out would not result in a sufficiently low crash severity. Consequently, areas away from paved and / or preceding traffic routes can be included for the determination of the emergency trajectory. In particular, off-road trajectories with a low risk of collision or trajectories with the smallest possible crash severity are preferred. In addition, depending on the dangerous situation, it may be necessary to bring the robot to a standstill as quickly as possible, for example in order to avoid driving into unrecognized areas in which living beings could potentially be. The robot can be decelerated in these cases, for example, by driving over rough surfaces such as meadows, fields, muddy areas or the like. In addition, collisions with objects such as bushes, shrubs and possibly trees can be approached to delay the robot.

In a further embodiment of the method, the object and / or the area that can be driven on by means of the robot lies outside a roadway that is normally provided for the robot.

This embodiment of the invention offers the advantage that the number of possible emergency trajectories is increased, since the space in which these are determined is increased. In addition, these areas in particular offer the potential to determine emergency trajectories that are further away from other road users. In these areas there are also a wide variety of objects that are not designed to be controlled by robots. This opens up new and previously untapped potential for optimizing the movements of a robot in emergency situations.

In a further embodiment of the method, the surroundings of the robot are divided into partial areas, the partial areas being assigned a partial area delay value based on at least one determination of the delay value, the determination of the emergency trajectory based on the sub-range delay values.

For efficient trajectory planning, it is advisable to divide and classify the areas that can be traveled by the robot, or to assign a partial area delay value to them. If an emergency trajectory is laid through such an area, it can be determined very efficiently and quickly what delay the robot has when driving in or through this area. If the robot is not decelerated to a standstill in one area, the sub-area deceleration value of the adjacent area can also be taken into account, and corresponding other areas. As a result, numerous possible emergency trajectories and their effects on the robot can be calculated in a few milliseconds, which enables a speed-optimized determination of an emergency trajectory.

Different variables can be used to determine the delay value. In one embodiment of the invention, the determination is based on a mass, size and / or rigidity of an object and / or a subsurface condition of an area. By taking these variables into account, the accuracy of a prediction for the deceleration of the robot can be increased. In particular, the set of impulses or methods from multi-body dynamics can be used to determine the momentum. Simulations carried out internally or externally can also be used. If this involves time-consuming simulations, only their input parameters and results can be used for the determination.

In a further embodiment of the method, this includes the additional step of determining whether the automated robot is occupied with living beings and / or dangerous goods. The determination of the emergency trajectory is based on the occupancy determined.

If the robot is, for example, a vehicle, an occupation can be understood to mean a number of people, their positions, sitting postures, their ages and the like. If the robot also transports cargo, its content, position, mass, hazard potential and the like can be understood as occupation.

This embodiment of the invention offers the advantage that the emergency trajectory can be determined as a function of the situation. Depending on the occupation of the robot, as explained above, different emergency trajectories and / or maximum accelerations of the robot come into question. The selection of the emergency trajectory is optimized by the step of determining the occupation.

In a further embodiment of the method, this includes the additional step of determining a probability of a collision with a person and / or another living being, the emergency trajectory being determined based on the determined probability.

For the determination of the probability, in particular, environmental data and / or the data of the robot, such as its speed, acceleration potential, and / or mass, can be used. Of particular interest here are the positions of people and / or living beings and, if applicable, their predicted courses of movement. Different options for determining the likelihood of a collision can be used for both the movements of the living beings and the robot.

This embodiment of the method offers the advantage that the emergency trajectory is based on an estimated real danger. In this way, unnecessary damage to the robot can be avoided, while at the same time optimizing the safety of living beings.

In a further embodiment of the method, a trajectory with a high, in particular the greatest possible, delay potential is selected as the emergency trajectory if the robot is not transporting any living beings and / or dangerous goods and a predefined probability of a collision with a person and / or another living being along this trajectory Exceeds threshold value, the delay potential of the trajectory being determined based on at least one determined delay value for a region or an object along the trajectory.

In a further embodiment of the method, a trajectory with the greatest possible delay potential that is below a predefined maximum threshold value is selected as the emergency trajectory if the robot transports living beings and / or dangerous goods and a probability of a collision with a person and / or another living being along this trajectory exceeds a predefined threshold value, the delay potential of the trajectory being determined based on at least one determined delay value for a region or an object along the trajectory.

In a further embodiment of the method, this includes the additional step of outputting a control signal for controlling the robot along the emergency trajectory. As a result, the robot controls and / or traverses the emergency trajectory.

The method can be carried out continuously in the robot in order to have an emergency trajectory ready at any time in the event of an emergency situation in which it is important to minimize damage to living beings and / or dangerous goods.

Continuous does not necessarily mean that the check must take place at the same time intervals. However, maximum times can be specified in which a new determination of an emergency trajectory is to take place. It is also possible that emergency trajectories are only calculated in specific driving situations. For example, whenever other road users have been detected or a probability for the presence of other road users exceeds a predefined threshold value.

This embodiment of the invention offers the advantage that safe operation of the robot can be ensured.

The approach presented here also creates a device which is designed to carry out, control or implement the steps of a variant of a method presented here in corresponding devices. The object on which the invention is based can also be achieved quickly and efficiently by means of this embodiment variant of the invention in the form of a device.

For this purpose, the device can have at least one processing unit for processing signals or data, at least one storage unit for storing signals or data, at least one interface to a sensor or an actuator for reading in sensor signals from the sensor or for outputting data or control signals to the Have an actuator and / or at least one communication interface for reading in or outputting data, which are embedded in a communication protocol. The computing unit can be, for example, a signal processor, a microcontroller or the like, wherein the storage unit can be a flash memory, an EEPROM or a magnetic storage unit. The communication interface can be designed to read in or output data wirelessly and / or wired, a communication interface that can read in or output wired data, for example, can read this data electrically or optically from a corresponding data transmission line or output it into a corresponding data transmission line.

In the present case, a device can be understood to mean an electrical device that processes sensor signals and outputs control and / or data signals as a function thereof. The device can have an interface which can be designed in terms of hardware and / or software. In the case of a hardware design, the interfaces can, for example, be part of a so-called system ASIC which contains a wide variety of functions of the device. However, it is also possible that the interfaces are separate, integrated circuits or at least partially consist of discrete components. In the case of a software-based design, the interfaces can be software modules that are present, for example, on a microcontroller alongside other software modules.

A computer program with program code, which can be stored on a machine-readable carrier or storage medium such as a semiconductor memory, a hard disk memory or an optical memory and is used to carry out, implement and / or control the steps of the method according to one of the embodiments described above, is also advantageous especially if the program product or program is executed on a computer or device.

A machine-readable storage medium is also claimed on which this computer program is stored.

Figure list

• 1 shows a schematic process diagram.

Working examples

In a first exemplary embodiment, the robot is an automatically operated motor vehicle for transporting people. The vehicle is equipped with several environmental sensors, including cameras, lidar sensors, radar sensors, ultrasonic sensors and microphones. The vehicle has several control units in which data from the environment sensors are processed and which can control the vehicle, at least in part, without human intervention in traffic. In order to increase the safety of vehicle occupants and other road users, the in 1 presented method for determining an emergency trajectory. The aim of the method is to provide the best possible emergency trajectory in situations in which a crash or a departure from a lane appears inevitable, resulting in the lowest possible crash severity and / or the lowest possible impact on living beings.

The procedure starts in step 101 .

In step 102 environmental data are received from the vehicle. On the one hand Receive sensor data from the vehicle's own sensors, as well as additional data from an external server via a wireless interface. In this exemplary embodiment, this is map data that already includes some information about objects located in the vicinity of the vehicle. Information such as delineator posts, signs, traffic lights and trees are received.

In step 103 an object and / or an area that can be traveled by the robot is detected based on the received environment data. For this purpose, there are classification algorithms that have been trained using machine learning processes

In step 104 a deceleration value is determined for the detected object and / or the area that can be driven over by the robot. In this exemplary embodiment, use is made of a database which assigns corresponding delay values to different objects and substrates. For objects that cannot be assigned to an object in the database, a delay value is estimated in this exemplary embodiment on the basis of the size of the object and the size of an anchoring point of the object (for example on the floor or a wall).

In step 105 an emergency trajectory through the determined object and / or along the area that can be driven by the robot is determined based on the determined deceleration value. In this exemplary embodiment, depending on the current traffic situation or dangerous situation, a maximum distance after which the robot should have come to a standstill is an absolute position at which the robot should come to a stop, an area which the robot should no longer be able to reach or a desired delay of the robot is specified. Based on this specification, the emergency trajectory is determined by the detected object, several detected objects or the area or areas detected by the robot.

The process ends in step 106 .

In a further example, the method takes place in a truck that transports people and also dangerous goods, in this case kerosene. By evaluating the sensor data recorded by means of an environment sensor system of the truck, a critical area was determined which the truck should not drive into under any circumstances, since there are pedestrians in the area. However, deceleration on the road intended for the truck cannot provide sufficient deceleration. In order to protect the people in the vehicle and the dangerous goods, a maximum deceleration value for the truck, which must not be exceeded, is specified in the method before the determination of the emergency trajectory. A trajectory that would lead through a solid tree, as a result of which the maximum deceleration value would presumably be exceeded, is consequently not considered as an emergency trajectory in this exemplary embodiment. Instead, in this exemplary embodiment, an emergency trajectory is determined that steers the truck through a crash barrier into a field so that the truck is decelerated by the crash with the crash barrier and by driving through the field and ultimately comes to a standstill without entering the critical area.

In a further exemplary embodiment, an automatically operated car transports neither dangerous goods nor living beings. In this vehicle, too, as in the previous example, an area is identified which the vehicle is not allowed to drive into. Due to the high speed of the vehicle, it is no longer possible to brake on the road intended for the vehicle. A collision with other vehicles on the regular traffic route is also out of the question as an emergency trajectory, as the ambient signals received cannot be used to rule out whether there are people or dangerous goods in these vehicles. Based on the received environment data, a tree with a trunk diameter of more than 40 cm is recorded in this exemplary embodiment. A comparison with a database is used to determine its deceleration value, which is sufficiently large to bring the vehicle to a complete standstill. The emergency trajectory is determined based on this deceleration value and the specification to bring the vehicle to a standstill as quickly as possible in such a way that the vehicle collides with the tree. In order to avoid possible injury to other living beings, a total loss of the vehicle is accepted.

In a further exemplary embodiment, the surroundings of the robot are divided into individual areas (“grid map”), with an area delay value being assigned to each area during the course of the method for determining an emergency trajectory. This allows the emergency trajectory to be determined quickly and efficiently using the range delay values alone. In this embodiment, the range delay value is not specified exactly, but in three different degrees, high, medium and low delay. In addition, a collision severity value is specified, no collision, low, medium or high collision severity. Both the area delay values and the collision severity values are determined based on the detected objects and / or areas that the robot can drive into. The calculation takes place in this Embodiment in the vehicle, whereby there is also the option of having this calculation carried out by an external computer and only receiving a "grid map" with corresponding values.

In this exemplary embodiment, the cargo of the robot is also classified. Indoor sensors such as cameras and weight mats are used for this. The classes "cargo worthy of protection", persons, dangerous goods, other living beings, as well as the class "cargo not worth protecting", no living beings or dangerous goods are introduced.

• - kleinstes Kollisionsrisiko, falls die gesamte Umgebung einsehbar ist und keine Gefahr für Lebewesen, insbesondere Menschen besteht
• - kleinste Kollisionsschwere, falls die gesamte Umgebung einsehbar ist und keine Gefahr für Lebewesen, insbesondere Menschen besteht, eine Kollision jedoch unausweichlich ist
• - maximale Geschwindigkeitsreduktion ohne Kollision, falls die gesamte Umgebung einsehbar ist und eine Gefahr für Lebewesen, insbesondere Menschen besteht, und der Roboter Personen oder Gefahrengüter transportiert
• - maximale Geschwindigkeitsreduktion ohne Kollision, falls nicht die gesamte Umgebung einsehbar und der Roboter Personen oder Gefahrengüter transportiert
• - maximale Geschwindigkeitsreduktion mit Kollision, falls die gesamte Umgebung einsehbar ist und eine Gefahr für Lebewesen, insbesondere Menschen besteht, und der Roboter keine Personen oder Gefahrengüter transportiert
• - maximale Geschwindigkeitsreduktion mit Kollision, falls nicht die gesamte Umgebung einsehbar ist und der Roboter keine Personen oder Gefahrengüter transportiert.

The determination of the emergency trajectory is based on an assessment of the traffic situation. Different criteria are specified for the choice of the optimized emergency trajectory, including:

• - Smallest risk of collision if the entire environment can be seen and there is no danger to living beings, especially people
• - Smallest collision severity, if the entire environment can be seen and there is no danger to living beings, especially people, but a collision is inevitable
• - Maximum speed reduction without collision, if the entire environment can be seen and there is a danger to living beings, especially people, and the robot is transporting people or dangerous goods
• - Maximum speed reduction without collision, if the entire environment cannot be seen and the robot is transporting people or dangerous goods
• - Maximum speed reduction with collision, if the entire environment can be seen and there is a danger to living beings, especially people, and the robot is not transporting any people or dangerous goods
• - Maximum speed reduction with collision if the entire area cannot be seen and the robot is not transporting any people or dangerous goods.

In this exemplary embodiment, the method ends with the output of a signal for controlling the robot along the specific emergency trajectory.

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• DE 102016211587 A1 [0002]

Claims (13)

Method for determining an emergency trajectory for an automated robot, comprising the steps: • Receiving (102) of environment data; • Detecting (103) at least one object and / or an area that can be traveled by the robot based on the surroundings data; • Determination (104) of a deceleration value for the detected object and / or the area that can be traveled by the robot; • Determination (105) of an emergency trajectory through the determined object and / or along the area that can be driven by the robot based on the determined deceleration value. Procedure according to Claim 1 , wherein the object and / or the area that can be driven by the robot lies outside a roadway that is normally provided for the robot. Method according to one of the preceding claims, wherein an environment of the robot is divided into partial areas and the partial areas are assigned a partial area delay value based on at least one determination of the delay value, the emergency trajectory being determined based on the partial area delay values. Method according to one of the preceding claims, wherein the deceleration value is determined based on a mass, size and / or rigidity of an object and / or a subsurface condition of an area. Method according to one of the preceding claims, with the step of determining an occupation of the automated robot with living beings and / or dangerous goods, wherein the determination of the emergency trajectory takes place based on the determined occupation. Method according to one of the preceding claims, with the step of determining a probability of a collision with a person and / or another living being, the emergency trajectory being determined based on the determined probability. Procedure according to Claim 5 and 6th , a trajectory with a high, in particular the greatest possible, delay potential being selected as the emergency trajectory if the robot is not transporting living beings and / or dangerous goods and a probability of a collision with a person and / or another living being along this trajectory exceeds a predefined threshold value, with the delay potential of the trajectory is determined based on at least one determined delay value for a region or an object along the trajectory. Procedure according to Claim 5 and 6th , whereby a trajectory with the greatest possible but below a predefined maximum threshold value, delay potential is selected as the emergency trajectory if the robot transports living beings and / or dangerous goods and a probability of a collision with a person and / or another living being along this trajectory exceeds a predefined threshold value , wherein the delay potential of the trajectory is determined based on at least one determined delay value for a region or an object along the trajectory. Method according to one of the preceding claims, with the additional step of outputting a control signal for controlling the robot along the emergency trajectory. Method according to one of the preceding claims, wherein the robot is designed as an industrial robot, lawn mower, watercraft, rail vehicle, motor vehicle, bus, shuttle, driverless vehicle, airplane, drone and / or helicopter. Device which is set up to carry out a method according to one of the preceding claims. Computer program, comprising instructions which cause the computer program to be executed by a computer, a method according to one of the Claims 1 until 10 to execute. Machine-readable storage medium on which the computer program is based Claim 12 is stored.

Images