Search

DE-102024133044-A1 - System for managing a vehicle fleet

DE102024133044A1DE 102024133044 A1DE102024133044 A1DE 102024133044A1DE-102024133044-A1

Abstract

The invention relates to a system for controlling a fleet (1) of autonomous vehicles (3) in road traffic, each autonomous fleet vehicle (3) being integrated into a control loop (R) together with a central control unit (5), in which the central control unit (5) is in mobile data communication with the fleet vehicles. According to the invention, the central control unit (5), as an AI-based unit, issues driving operation instructions (y) to the individual autonomous fleet vehicles (3) based on actual sensor data (x) from the fleet vehicles (3) and on the basis of a central, vehicle-wide fleet target value (z).

Inventors

  • Dirk Bäder
  • Michael Grabowski

Assignees

  • AUDI AKTIENGESELLSCHAFT

Dates

Publication Date
20260513
Application Date
20241112

Claims (10)

  1. System for controlling a fleet (1) of autonomous vehicles (3) in road traffic, each autonomous fleet vehicle (3) being integrated into a control loop (R) together with a central control center (5), in which the central control center (5) is in mobile data connection with the fleet vehicles, characterized in that the central control center (5) as an AI-based unit issues driving operation instructions (y) to the individual autonomous fleet vehicles (3) based on actual sensor data (x) of the fleet vehicles (3) and on a central, vehicle-spanning fleet target specification (z), so that, in particular, central fleet parameters can be optimized with regard to the fleet target specification (z) by coordinating the fleet vehicles (3).
  2. System according to Claim 1 , characterized in that the central fleet target specification (z) relates to the following matters: - optimization of the total travel time of all autonomous fleet vehicles (3); - reduction of the total fuel consumption of all autonomous fleet vehicles (3); - grouping of autonomous fleet vehicles (3) of the same or similar performance class on a common lane (v1, v2); - rapid resolution of a traffic jam in which the fleet vehicles (3) are located; - real-time modulation of the driving speeds of the fleet vehicles (3); - elimination of a traffic control system consisting of traffic lights, fixed lanes, traffic signs or the like; and/or - coordinated driving of several fleet vehicles (3) with a minimum distance (a) on a limited roadway area, for example, driving of three fleet vehicles (3) side by side on a two-lane highway.
  3. System according to Claim 1 or 2 , characterized in that the control loop (R) incorporates both vehicle-internal sensors, such as camera, radar, lidar or GPS, and vehicle-external sensors, such as camera, temperature sensor, humidity sensor.
  4. System according to one of the preceding claims, characterized in that the fleet target specification (z) relates to the setting of small vehicle distances (a) to optimize the overall Cd value of the fleet (1), in particular taking into account the vehicle external geometries and/or the individual Cd values of the fleet vehicles (3).
  5. System according to one of the preceding claims, characterized in that the fleet target specification (z) relates to a grouping of the fleet vehicles (3) onto specific lanes (v1, v2), and in particular that the grouping is carried out depending on vehicle-specific characteristics of the fleet vehicles (3), namely, for example, performance class or drag coefficients of the fleet vehicles (3) or the like.
  6. System according to one of the preceding claims, characterized in that the central control center (5) detects non-autonomous or non-communicating road users (9), in particular by means of the sensors (7) of the fleet vehicles, and in particular that the central control center (5) issues the driving operation instructions (y) to the fleet vehicles (3) taking into account the road users (9).
  7. System according to one of the preceding claims, characterized in that the central control center (5) recognizes at least one priority special vehicle (11), in particular by means of the sensors (7) of the fleet vehicles (7), and in particular that the central control center (5) takes into account the priority special vehicle (11) and issues driving operation instructions (y) to the fleet vehicles (3) in order to enable prioritized advancement of the priority special vehicle (11), for example driving operation instructions (y) to form a lane for the priority special vehicle (11).
  8. System according to one of the preceding claims, characterized in that the central control center (5) detects a short-term road closure, for example due to an accident, in particular by means of the sensors (7) of the fleet vehicles (3), and/or that in the event of such a road closure the central control center (5) issues driving operation instructions (y) to the fleet vehicles (3) in order to enable a dynamic autonomous or AI-supported and moderated rerouting of traffic, for example to the opposite lane, so that barriers and/or road guidance systems are required.
  9. System according to one of the preceding claims, characterized in that the central control center (5) detects an open, unregulated traffic area (13) without a traffic-structuring traffic control system, in particular by means of the sensors (7) of the fleet vehicles (3), and in particular that the central control center (5) issues driving operation instructions (y) to the fleet vehicles (3) entering the traffic area (13) in order to enable smooth traffic flow through the traffic area.
  10. System according to Claim 9 , characterized in that the central control center (5) detects non-autonomous or non-communicating road users (9) in the unregulated traffic area (11), in particular by means of the sensors (7) of the fleet vehicles (3), and in particular that the central control center (5) issues the driving operation instructions (y) taking these road users (9) into account.

Description

The invention relates to a system for controlling a vehicle fleet according to the preamble of claim 1. In the field of vehicle automation, the vehicle uses onboard sensors such as cameras, radar, and lidar to autonomously optimize driving parameters like speed, lane selection, and distances to other road users. These systems are designed to make driving decisions based on this sensor data, pursuing their own objectives, such as the shortest travel time or minimum energy consumption. Advanced driver assistance systems are also common in this context, capable of coordinating driving maneuvers and maintaining safe distances to a limited extent. Currently, such autonomous vehicles operate largely in isolation. This means they optimize their driving parameters independently and are primarily focused on their own goals and surroundings. A higher-level, fleet-wide coordination system is lacking, which optimizes key parameters such as overall energy efficiency or the total travel time of a vehicle fleet. This isolated optimization often leads to inefficient behavior in traffic: road space is not used optimally. This lack of coordination can result in longer traffic jams and higher energy consumption. Furthermore, in situations where autonomous vehicles encounter non-autonomous road users or operate in complex traffic areas without clearly defined rules, their driving style is often inconsistent, which can negatively impact traffic flow and safety. From the DE 10 2020 130 387 A1 A system for planning vehicle routes using a modified Nash equilibrium solution for a multi-agent game is known. The system uses adaptive search optimization to determine optimal routes for autonomous vehicles based on rewards and penalties for various vehicle actions. This methodology allows for the simulation and prediction of future traffic situations and considers the interactions of various agents such as vehicles, pedestrians, and static objects. The goal is to calculate the safest and most efficient route for each vehicle, especially in complex traffic scenarios, and to transmit this route to the vehicles via a central control unit. From the DE 10 2023 123 008 A1 A vehicle guidance system is known that requests additional support from a server in certain driving situations. For example, if the autonomous vehicle encounters an unforeseen situation such as a blocked lane, it sends a request to the server, which then provides alternative routes. These routes consist of waypoints that guide the vehicle around the obstacle. The server analyzes the vehicle's sensor data and provides this guidance information in real time. The system is designed to facilitate the handling of critical driving situations and ensure that the vehicle follows the optimal route despite obstacles. From the DE 10 2017 217 444 A1 is a method and system for the continuous updating of control models for autonomous vehicles. A central control unit creates data collection tasks that define specific conditions for data gathering. The mobile units then collect sensor data and transmit it back to the central unit, which updates the control model based on this data. The system enables flexible adaptation of the model to current environmental conditions, thereby optimizing the vehicles' driving characteristics in real time. This approach offers continuous control adjustment and responds flexibly to changing traffic conditions. The object of the invention is to provide a system that enables safe, resource-efficient and/or smooth traffic management for a fleet of autonomous vehicles. The problem is solved by the features of claim 1. Preferred embodiments of the invention are disclosed in the dependent claims. The invention relates to a system for controlling a fleet of autonomous vehicles in road traffic. Each of the autonomous fleet vehicles is integrated into a control loop together with a central control center. Within this control loop, the central control center maintains a mobile data connection with all fleet vehicles. According to the characterizing part of claim 1, the following measures are taken to achieve safe, resource-efficient, and/or smooth traffic flow for the fleet vehicles: The central control center, designed as an AI-based unit, issues driving operation instructions to the individual autonomous fleet vehicles based on actual sensor data from the fleet vehicles and a central, vehicle-wide fleet target value. In this way, central fleet parameters can be optimized with regard to the fleet target value by coordinating the fleet vehicles. According to the invention, the fleet vehicles are centrally controlled by AI, which Coordinated driving operations enable efficient optimization of fleet parameters. In contrast to conventional systems, which often rely on vehicle-specific optimizations, this approach offers holistic optimization at the fleet level, leading to more efficient traffic flows and better resource allocation. In a technical implementation, the cent