Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

How to Drive -- An Ability-based Description of Autonomous, Remote and Human Driving

Florian Pfab, Nils Gehrke, Frank Diermeyer

TL;DR

This paper tackles the lack of a precise, holistic description of the abilities required for safe operation of driving systems in public traffic. It proposes a four-step method to synthesize a holistic, solution-neutral ability graph by converting and merging driving-task descriptions from multiple sources, then refines usability for practical use. The resulting graph covers autonomous, remote, and human-in-the-loop driving, enabling requirement validation, test design, and online monitoring, and is demonstrated against a German driving exam task, an open-source AV stack, and teleoperation scenarios. The work provides a normative framework to reason about capability coverage, fault detection, and responsibility allocation across operators and systems, with clear paths for extension as driving technologies evolve.

Abstract

The development of autonomous and remote-operated driving systems requires extensive stakeholder analyses, requirement engineering, and formalized system descriptions. This is necessary to guarantee the success of the final product after the expensive and time-consuming development phase. To integrate a formalized description of the required abilites of the system, ability graphs have been proposed in the literature. Up to this date, however, this ability graph has only been used to model less complicated driver assistance systems in the literature. This work aims to introduce the value of an ability graph-based description of complex driving systems. This is achieved by successfully demonstrating and discussing a method for constructing a holistic ability graph capable of describing the entirety of abilities required for any driving system.

How to Drive -- An Ability-based Description of Autonomous, Remote and Human Driving

TL;DR

This paper tackles the lack of a precise, holistic description of the abilities required for safe operation of driving systems in public traffic. It proposes a four-step method to synthesize a holistic, solution-neutral ability graph by converting and merging driving-task descriptions from multiple sources, then refines usability for practical use. The resulting graph covers autonomous, remote, and human-in-the-loop driving, enabling requirement validation, test design, and online monitoring, and is demonstrated against a German driving exam task, an open-source AV stack, and teleoperation scenarios. The work provides a normative framework to reason about capability coverage, fault detection, and responsibility allocation across operators and systems, with clear paths for extension as driving technologies evolve.

Abstract

The development of autonomous and remote-operated driving systems requires extensive stakeholder analyses, requirement engineering, and formalized system descriptions. This is necessary to guarantee the success of the final product after the expensive and time-consuming development phase. To integrate a formalized description of the required abilites of the system, ability graphs have been proposed in the literature. Up to this date, however, this ability graph has only been used to model less complicated driver assistance systems in the literature. This work aims to introduce the value of an ability graph-based description of complex driving systems. This is achieved by successfully demonstrating and discussing a method for constructing a holistic ability graph capable of describing the entirety of abilities required for any driving system.
Paper Structure (22 sections, 5 figures)

This paper contains 22 sections, 5 figures.

Table of Contents

  1. INTRODUCTION
  2. Potential Application Areas
  3. Contributions
  4. RELATED WORK
  5. Ability and skill graphs for av
  6. Vehicle driving tasks
  7. Driver-vehicle-environment system model
  8. Driving task allocation
  9. Requirements from the vehicle guidance task
  10. ddt
  11. Three-level hierarchy of the driving task
  12. AV
  13. THE ABILITY-GRAPH DESCRIPTION FOR DRIVING
  14. Graph deduction of existing task descriptions (Step 1)
  15. Transformation into ability graphs (Step 2)
  16. ...and 7 more sections

Figures (5)

  • Figure 1: Examplary ability graph for the road surface detection task.
  • Figure 2: Representation of the method used for constructing the holistic ability graph.
  • Figure 3: The node and edge count change during the different stages from the original driving task literature towards the final holistic ability graph as presented in \ref{['sec:abilitygraph']}.
  • Figure 4: The final holistic ability graph. To reduce the complexity in this figure, the level of sub-abilities is reduced to four. More detail can be obtained by visualizing the graph using the tool in LinkGraph.
  • Figure 5: The different applications and methods used to validate the derived ability graph.