Method for remotely controlling multiple driverless self-driving systems, as well as a control center for remotely controlling the self-driving systems and system

Abstract

Method for remotely controlling several driverless self-driving systems (12) for transporting persons and / or goods, wherein a central, stationary control station (11) is provided which has several control resources (16) and from which each of the self-driving systems (12) is remotely controlled by means of one of the control resources (16) in the event of a respective assistance requirement of the self-driving system (12), characterized in that - at least a future time interval (22) for which one of the self-driving systems (12) requires assistance, and a corresponding location (21) where the need for assistance exists, are determined, - to which at least one time interval (22) and / or to which the respective location (21) is assigned one of the control resources (16) of the control room (11) in an operating plan (17) and - each control resource (16) in the time interval (22) to which it is allocated according to the operating plan (17) is coupled to the self-driving system (12) requiring assistance in the time interval (22) and / or for each location (21) to which it is allocated according to the operating plan (17), successively to at least some of the self-driving systems (12) requiring assistance at the location (21) for remote control of the same, wherein at least two at least partially overlapping time intervals (22) are each detected for one of the self-driving systems (12) and a respective control command (24) for setting a displacement measure is sent to at least one of the self-driving systems (12) before the time interval (22) for which the self-driving system (12) requires assistance, wherein the displacement measure is configured to shift the time interval (22) by comprising the displacement measure,that a departure of the self-driving system (12) is delayed and / or a driving speed is changed, thereby eliminating the overlap.

Description

[0001] The invention relates to a method for coordinating several driverless self-driving systems. Such a self-driving system can be designed for transporting people and / or goods. The self-driving system operates driverless by means of an autopilot; that is, all persons are passengers. The invention also includes a control station, or simply a control center, by means of which one of the self-driving systems is remotely controlled if, for example, for safety reasons in a complex traffic situation, it needs to be driven by a human driver due to the autopilot being overwhelmed. Finally, the invention also includes a system consisting of the control center and a fleet of self-driving systems.

[0002] In connection with the invention, a motor vehicle that drives itself automatically without a driver, using an autopilot, is referred to as a self-driving system (SDS). A self-driving system can be used for transporting people as well as goods and merchandise. Self-driving systems are equipped with onboard sensors for detecting their surroundings. Such sensors can include, for example, at least one camera and / or at least one laser scanner and / or at least one radar, to name just a few examples. A self-driving system can provide the capability to communicate with other road users and / or infrastructure (for example, a server, roadside infrastructure such as a traffic light, and / or the aforementioned control center). Another term for such a control center is hand center.

[0003] However, situations can arise in which a self-driving system reaches the limits of its capabilities. For example, a self-driving system may be unable to adequately assess or oversee the traffic situation based on its own sensors and / or information from infrastructure and other road users, or it may be unable to plan a trajectory. In such a situation, the self-driving system requires assistance. A control center can support the self-driving system by remotely controlling it from the control center, for example, through a human operator and / or a computer system, by sending control commands from the control center to the self-driving system.

[0004] The described concept is known, for example, from DE 10 2014 014 119 A1. A self-driving system that sends a request for assistance to a control center and is then remotely controlled is also known from US 9 465 388 B1.

[0005] From DE 10 2016 001 264 A1 it is known that a self-driving system can also be remotely controlled by an external control unit, so that no human operator is required in the case of remote control.

[0006] From DE 10 2015 118 489 A1, a self-driving system is known that detects an unexpected driving environment and then sends a request for assistance via remote control together with sensor data to a control center.

[0007] If a fleet of autonomous vehicles is operated by a company, such as a taxi company, then an operator must be on standby in the control room in case a vehicle requires assistance. If several vehicles require assistance simultaneously, a corresponding number of operators must be on standby. To minimize personnel requirements, or, in the case of computer-aided remote control, computing resources, the autonomous vehicles must be coordinated so that their assistance needs do not tie up too many resources in the control room.

[0008] German patent DE 20 2013 100 347 U1 describes a remote control system for motor vehicles in which a central server system receives remote control requests from motor vehicles and then connects the respective motor vehicle to a personal data terminal, which performs the remote control of the motor vehicle in the manner of a driving simulation computer game. The remote control includes the control of the longitudinal and lateral speed and acceleration of the motor vehicle.

[0009] The invention is based on the objective of efficiently operating a control center for self-driving systems.

[0010] The problem is solved by the subject matter of the independent patent claims. Advantageous embodiments of the invention are described by the dependent patent claims, the following description, and the figures.

[0011] The invention provides a method for remotely controlling multiple driverless self-driving systems designed or equipped for transporting persons and / or goods. These self-driving systems thus form a fleet. Transport can take place on a road network. A central, stationary control station is provided for remote control, which has several control resources. Each control resource can remotely control one self-driving system. From the control station, each self-driving system is remotely controlled by one of the control resources whenever assistance is required. Therefore, as already described in the introduction, the self-driving systems are only partially autonomous.

[0012] To avoid having to provide an arbitrarily large number of control resources—in the worst case, one control resource for each of the self-driving systems—the method according to the invention comprises the following measures. At least one future time interval, during which each of the self-driving systems will require assistance, and a corresponding location where this assistance is required are determined. One of the control resources of the control center is allocated to each of these time intervals in an operating plan. Thus, whenever a control resource becomes available because a time interval has ended, this control resource can be used again in another time interval. In addition to or as an alternative to allocating the control resources according to time intervals, it can be provided that one of the control resources of the control center is allocated to each location where assistance is required in the operating plan.Each control resource is then coupled with the self-driving system requiring assistance during the time interval to which it is allocated according to the operating plan. In the second variant, each control resource can be coupled sequentially with at least some of the self-driving systems requiring assistance at each location to which it is allocated according to the operating plan. This coupling is performed for the purpose of remotely controlling the respective self-driving system. Thus, the described control commands can be sent or transmitted from the control resource to the self-driving system, enabling the control center's control resource to remotely control the respective self-driving system. This allows the control resource to guide the self-driving system, i.e., to perform longitudinal control (acceleration and braking) and / or lateral control (steering).

[0013] The inventive method thus assigns control resources of a control center to a self-driving system, a road segment, or a traffic situation, with the latter two variants being combined in the description as a single location where the road segment is situated or the traffic situation occurs. The method therefore enables a control resource to be used for different self-driving systems; specifically, at the end of each time interval, the control resource can be assigned to a different time interval, provided the two time intervals do not overlap. This allows a control resource to be efficiently scheduled continuously within an operating plan.

[0014] To further optimize or increase the efficiency of the control center's operation, it is planned that at least two partially overlapping time intervals will be detected, each representing one of the autonomous driving systems. Overlapping time intervals indicate that two autonomous driving systems simultaneously require assistance, since each time interval represents one autonomous driving system and overlapping time intervals thus suggest two autonomous driving systems requiring simultaneous remote control. This ties up two control resources. To avoid this, and thus enable both autonomous driving systems to be remotely controlled with a single control resource, it is planned that a control command will be sent to at least one of the autonomous driving systems before the time interval in which it requires assistance. This command will initiate a shift action. This shift action is designed to postpone the time interval.This eliminates the overlap.

[0015] The rescheduling measure involves delaying the departure of the autonomous driving system. The system therefore departs later, resulting in a different, delayed, or postponed arrival time at the location requiring assistance. Additionally or alternatively, the rescheduling measure involves changing the speed of the autonomous driving system. This, too, can shift or change the arrival time at a location requiring assistance. Increasing the speed can bring the arrival time forward, while decreasing the speed can push it back to a later time.

[0016] The invention also includes further developments that result in additional advantages.

[0017] The operating plan should be flexibly adapted to changing traffic situations. For example, the time interval of an autonomous driving system may shift if the system reaches the traffic situation or the location where assistance is needed earlier or later than originally planned. Accordingly, updated route data is preferably received repeatedly from one, several, or all of the autonomous driving systems, and the operating plan is cyclically adjusted to the updated route data.

[0018] The time intervals at which each of the autonomous driving systems requires assistance can be determined by the control center itself (i.e., a server or computer) or by the respective autonomous driving system itself. To determine the specific assistance requirement, each autonomous driving system can receive route data for a planned route from at least one of the other systems. This route data can include, for example, the current position and / or the planned trajectory. The assistance requirement of the autonomous driving system can then be determined on a digital map by a processor in the control center for that system. Additionally or alternatively, demand data directly reporting or describing the time interval and location for the planned assistance requirement can be received from at least one of the autonomous driving systems.This can be used if a self-driving system can independently determine its assistance needs.

[0019] The control station can provide at least one of the following as a control resource: a human operator or a group of human operators (if a self-driving system is to be remotely controlled by multiple operators) or the computing time of a control computer.

[0020] One question that arises is how to determine the need for assistance in advance. The specific need for assistance is identified, in particular, if a location with at least one recurring, predetermined traffic situation lies along the route of the respective autonomous driving system. Such a traffic situation could be, for example, a traffic jam, or a vehicle density or frequency exceeding a certain threshold. Additionally or alternatively, a need for assistance can be identified if a location with at least one predicted, predetermined traffic situation lies along the route of the autonomous driving system. For example, if it is known that a predetermined traffic situation, such as a traffic jam or a detour, will occur due to a predetermined event, such as a football match or a demonstration, then a need for assistance for an autonomous driving system can also be identified in this case.Assistance requirements can also be identified if at least one predetermined road segment lies along the route of the autonomous driving system. For example, a road segment with an above-average accident rate, as shown in accident statistics, can be designated as a location requiring assistance from autonomous driving systems. The corresponding time interval is defined as the expected dwell time of the autonomous driving system at the respective location. This dwell time can be determined based on the planned route and the speed set by the autopilot, or the planned speed. Additionally or alternatively, traffic flow data indicating average speeds along the route can be used.

[0021] Another measure for optimizing the operation of the control center involves sending a control command to one of the autonomous driving systems to form a platoon (column) with at least one other control system. In a platoon, one autonomous driving system follows another and thus adopts its driving behavior. Therefore, if the leading autonomous driving system of a platoon is remotely controlled, every subsequent autonomous driving system in the platoon automatically follows without needing to be remotely controlled itself. By forming the platoon, the time interval of the first autonomous driving system (which attaches itself to or follows another autonomous driving system in the platoon) is combined with the time interval of the second autonomous driving system, which leads the platoon, into a single time interval.This means that only one control resource is needed to remotely control two or more self-driving systems during this time interval, by remotely controlling only the leading self-driving system of the platoon and having each other self-driving system of the platoon follow it.

[0022] A control station operated according to the method of the invention constitutes one embodiment of the control station according to the invention. In other words, the invention provides a control station for the remote control of several driverless self-driving systems. Each self-driving system can transport at least one person and / or goods. The control station includes a processor unit configured to carry out one embodiment of the method of the invention. For this purpose, the processor unit can include at least one microprocessor and / or at least one microcontroller. An operating plan for the control resources of the control station can be generated using the processor unit. The control resources themselves can, of course, also be operators, as described. The processor unit can include program code configured to carry out the embodiment of the method of the invention.The program code can be stored in a data memory of the processor device.

[0023] As described, the invention also includes the system comprising the control station and several driverless self-driving systems. The self-driving systems are configured to operate in situations requiring assistance, where the self-driving system recognizes that autonomous driving is impossible, by responding to a remote control signal from the control station, i.e., remote control commands or control commands from the control station.

[0024] The following are exemplary embodiments of the invention described. This is illustrated by: Fig. 1 a schematic representation of an embodiment of the system according to the invention; Fig. 2. A sketch illustrating time intervals measured by a control station of the system of Fig. 1 can be determined; and Fig. 3 a sketch to illustrate the time intervals after they have been shifted relative to each other by at least one shift operation.

[0025] The embodiment described below is a preferred embodiment of the invention. In this embodiment, the described components each represent individual features of the invention that can be considered independently of one another. Each of these features further develops the invention independently and can therefore be considered part of the invention individually or in a combination other than that shown. Furthermore, the described embodiment can also be supplemented by other features of the invention already described.

[0026] In the figures, functionally identical elements are each provided with the same reference symbols.

[0027] Fig. Figure 1 shows a system 10, which can include a control station 11 and self-driving systems 12 or SDS 12. The control station 11 can include a processor unit 13, which can communicate with each of the SDS 12 via a respective communication link 14. The processor unit 13 can, for example, be connected to the internet 15. Each communication link 14 can then include, for example, an internet connection and / or a radio connection. The control station 12 can also include control resources 16 for remotely controlling the self-driving systems 12, if a self-driving system 12 requires assistance. An operating plan 17 can be automatically generated for the control resources 16 by means of the processor unit 13. The operating plan 17 can specify for each control resource 16 the time interval, driving situation, or location at which a self-driving system 12 is to be remotely controlled by the respective control resource 16.

[0028] To determine the operating plan 17, the control station 11 can receive demand data 18 and / or route data 19 from each self-driving system 12. Using the route data 19, a self-driving system 12 can signal its planned route 20 to the control station 11. If no demand data 18 is received from a self-driving system, the processor unit 13 can determine the assistance requirement of the self-driving system 12 itself based on the route data 19. For example, a digital road map 21 can be used to identify, based on the routes 20, whether a self-driving system 12 will pass a location 21 where assistance may be required. Furthermore, it can be determined within which time interval 22 the respective self-driving system 12 will be at the location 21. "Location" here refers to an area.

[0029] Fig. Figure 2 illustrates the resulting time intervals 22 for location 21 by way of example. The time intervals 22 are arranged over time t. Within an overlapping area 23, two control resources 16 are occupied or required for the respective overlapping time intervals 22.

[0030] In Fig. 2. The SDS 12 are distinguished by their respective designations SDS 1, SDS 2, SDS 3, SDS 4.

[0031] Through the processor device 13, a control command 24 can now be sent to each SDS 12 or to one or several SDS 12s (see Fig. 1) The respective control command 24 can trigger or control a displacement action in the respective SDS 12, as has already been described.

[0032] Fig. Figure 3 illustrates a possible result. As in Fig. As illustrated in Figure 3, the overlapping areas could be resolved or removed. Thus, instead of 3 control resources, as in the example of Fig. 2 was needed, only two tax resources in the Fig. Figure 3 illustrates the case as necessary.

[0033] The operating plan 17, as it is according to Fig. The result of 3 can then be output or displayed for the allocation of control resources 16. Each control resource allocated according to the operating plan for location 21 or an individually allocated SDS 12 can then send a remote control signal 25 to the respective SDS 12, for example by means of the processor unit 13, and thereby remotely control the respective SDS 12 and thus remotely control or guide it through location 21 or its area, so that the SDS 12 does not have to drive itself at location 21 using its autopilot.

[0034] A particularly preferred embodiment is described below. A method is performed which helps to ensure that each SDS in the managed SDS fleet has a commander (control resource) available if required, or that such a commander becomes available within a defined period.

[0035] The invention assumes that at least some of the situations in which a self-driving system (SDS) needs to request support from the command center are known and predictable. These can be recurring traffic situations as well as road sections with features that require assistance (e.g., bottlenecks, complex road layouts). They could also include road sections that an SDS can navigate independently under normal traffic conditions, but not under certain traffic situations whose occurrence is predictable (e.g., morning and evening rush hour).

[0036] The basic idea can be described as follows: • A central forecast is generated to determine when each SDS will require which type of assistance from a commander and for how long. Alternatively, a forecast could be generated by the SDS itself and made available to the command center for centralized processing. A combination of forecasts from the command center and the individual SDSs, followed by centralized evaluation, is also conceivable. • Based on this forecast, the temporal allocation of commanders to SDSs is planned. Commanders could also be assigned to specific road segments or traffic situations (e.g., a particular intersection or a section of road currently occupied by a garbage truck). For each road segment, a temporal allocation could then be made, resulting in a combination of spatial and temporal assignments. • The journey of an SDS can be influenced by the Command Center within certain limits to control the arrival time of an SDS at a location requiring assistance (e.g., by influencing travel speed, departure time, or route). The goal is to avoid overlapping assistance times whenever possible and to achieve a sequential response. • The movement of an SDS can be influenced within certain limits by the Command Center so that several supervised SDSs travel in succession. In this case, a spatial assignment of a commander would be useful, and assistance could be provided for SDSs traveling in succession or at short intervals (in time and space). • It would also be conceivable to virtually couple SDS vehicles traveling one behind the other (electronic drawbar), which could then be controlled as a unit by the Command Center and thus, for example, be guided as one unit across an intersection.

[0037] Possible input variables for the forecast are the following: • SDS: ◯ Current position ◯ Current speed ◯ Current status ◯ Planned route / remaining route of the SDS ◯ Optional: Time or distance to the next non-automatically navigable section of the route • Street (static input data): ◯ Road map with rated road segments (rating according to the performance of the SDS used). For traceability and verification, this rating should also include a justification (for example: "frequent pedestrians at the roadside", "narrow road layout", "complex traffic situation", ...) as well as the date of the last update of this information. • Traffic (dynamic input data): ◯ Historical traffic data • Current traffic information from the SDS fleet, swarm data from other vehicles, and other sources such as the traffic control center. This traffic information can also include local (possibly short-term) disruptions such as roadworks, mobile roadworks due to maintenance, garbage collection, street sweeping, etc. Optionally, the location and movement profiles of other road users via smartphone, activity tracker, or similar devices can also be considered, especially pedestrians and cyclists. ◯ Information about traffic control (traffic light information, shoulder opening, lane closure, etc.) from a traffic control center or a corresponding provider or service provider. • Predicted special traffic situations due to, for example, major events (concerts, sporting events). This prediction can be based on the evaluation of event calendars and / or social media.

[0038] The process may then include the following steps: Step 1 • For each SDS to be supervised ◯ Capturing the current position ◯ Forecast of the remaining journey (location and time profile) based on current traffic information and expected traffic development ◯ Forecast of times requiring assistance and type of assistance (start and end time, time interval) ◯ Updates at fixed time intervals (e.g., once per minute) Step 2 • Creating an assistance schedule for each SDS (plan with the time intervals required for each SDS, SDS-oriented commander deployment plan) • Alternatively or additionally, a schedule could be created for each section of road to be supervised (e.g., intersections etc. which require assistance from a commander, road- or traffic-situation-oriented deployment plan). • Evaluation of the assistance schedules for all SDS or road sections to be supervised ◯ Identification of parallel time intervals ◯ Formation of groups of parallel time intervals ◯ Assessment of the degree of temporal overlap (e.g., assignment to an assessment grid consisting of, for example, 15-second blocks) ◯ Identification of time intervals (with low temporal overlap) that have the potential to be shifted against each other by influencing the SDS (e.g. travel speed, departure time, adjustment of the route and thus the travel time and / or the need for assistance) in such a way that successive processing becomes possible. • Identifying the requirements for each commander, depending on the characteristics of the SDS and the difficulty of the assistance task to be performed. There might be commander tasks that are easy to handle, take little time, and perhaps can even be performed in parallel with other, simple commander activities. On the other hand, there might also be commander activities with high cognitive demands on the commander, which also require a specific qualification (for example, a special SDS license). Step 3 • Assignment of SDS or road sections to commanders • Check whether the predicted need for assistance can be met. ◯ Consideration of free commander capacities ◯ Consideration of a defined reserve for unforeseen assistance needs ◯ Identifying conflicts in planning ◯ Assessment of the conflicts that have arisen (Which SDS represent which sections and which times are in competition with each other?) Step 4 • Identification of approaches to conflict resolution ◯ Planning for additional commanders ◯ Interference in the planning of SDS departure times, travel speeds, waiting times, and routes ◯ Limitation of the number of SDS units used Step 5 • Sending control commands to the SDS to be influenced • Schedule update • If necessary, inform passengers: “There is a traffic jam… Your arrival time will be optimized…” • If necessary, inform other central facilities of mobility providers, traffic control centers, etc., if the intervention of a commander has not only purely local effects.

[0039] The process is repeated cyclically throughout the operation.

[0040] The procedure is also used for the general operational planning of the commanders, e.g. for planning additional capacities for special events, such as a major event.

[0041] Overall, the example shows how the invention can provide a scheduling or operating plan for a control center for self-driving systems (SDS - Self Driving System). Reference symbol list 10 System 11 Control Center 12 Self-driving systems 13 Processor setup 14 Communication link 15 Internet 16 tax resource 17 Operating plan 18 demand data 19 route data 20 Route 21 Place 22 Time interval 23 Overlap 24 Control command 25 Remote control signal

Claims

[1] Method for remotely controlling several driverless self-driving systems (12) for transporting persons and / or goods, wherein a central, stationary control station (11) is provided which has several control resources (16) and from which each of the self-driving systems (12) is remotely controlled by means of one of the control resources (16) when the self-driving system (12) requires assistance, characterized by , that - at least a future time interval (22) for which one of the self-driving systems (12) requires assistance, and a corresponding location (21) where the need for assistance exists, are determined, - to which at least one time interval (22) and / or to which the respective location (21) is assigned one of the control resources (16) of the control room (11) in an operating plan (17) and - each control resource (16) in the time interval (22) to which it is allocated according to the operating plan (17) is coupled to the self-driving system (12) requiring assistance in the time interval (22) and / or for each location (21) to which it is allocated according to the operating plan (17), successively to at least some of the self-driving systems (12) requiring assistance at the location (21) for remote control of the same, wherein at least two at least partially overlapping time intervals (22) are each detected for one of the self-driving systems (12) and a respective control command (24) for setting a displacement measure is sent to at least one of the self-driving systems (12) before the time interval (22) for which the self-driving system (12) requires assistance, wherein the displacement measure is configured to shift the time interval (22) by comprising the displacement measure,that a departure of the self-driving system (12) is delayed and / or a driving speed is changed, thereby eliminating the overlap. [2] Method according to claim 1, wherein repeatedly updated route data (19) are received from one or some or each of the self-driving systems (12) and the operating plan (17) is cyclically adapted to the updated route data (19). [3] Method according to one of the preceding claims, wherein each of the self-driving systems (12) is used to determine the respective assistance requirement a) route data (19) for each planned route (20) are received from at least one of the self-driving systems (12) and the assistance requirement of the self-driving system (12) is determined in a digital road map (MAP) by a processor unit (15) of the control room (11) and / or b) demand data (18) are received from at least one of the self-driving systems (12), which report the time interval (22) and the location (21) for the planned assistance requirement of the self-driving system (12). [4] Method according to any of the preceding claims, wherein at least one of the following is provided as the respective control resource (16): a human operator, a group of human operators, computing time of a control computer. [5] Method according to one of the preceding claims, wherein the respective assistance requirement is determined if along a route (20) of the respective self-driving system (12) there is a location (21) with at least one recurring predetermined traffic situation and / or at least one predicted predetermined traffic situation and / or at least one predetermined road segment, and the time interval (22) is defined as an expected dwell time of the self-driving system at the location. [6] Method according to one of the preceding claims, wherein a control command to form a platoon with a second of the self-driving systems (12) is sent to a first of the self-driving systems (12), wherein by forming the platoon the time interval (22) of the first self-driving system is combined with the time interval (22) of the second self-driving system (12) to form a single time interval. [7] Control station (11) for remotely controlling several driverless self-driving systems (12) for transporting persons and / or goods, wherein the control station (11) has a processor unit (15) configured to perform a method according to one of the preceding claims. [8] System (10) comprising a control station (11) according to claim 7 and several driverless self-driving systems (11) for transporting persons and / or goods, wherein the self-driving systems (12) are configured to drive in the event of a need for assistance in which the self-driving system (12) recognizes that self-driving is impossible, depending on a remote control signal (25) of the control station (11).

Images