Autonomous Vehicles Safety Analysis

Objectives:

The main objective of the project is to explore possible strategies for safety analysis and argumentation of autonomous mobile systems taking into account that the behaviour of such systems is not determined by algorithms which define responses for particular events and states, but is based on situation awareness model and decision making process.

Contact:

The contact for this research project is here.

Rationale:

Generally autonomy relates to freedom to determine one’s own actions and behaviour. Autonomy of a mobile system can be interpreted as the ability to accomplish a given mission without human intervention. That means that the system should be able to make decisions how the mission goals could be achieved and how to cope with changes in the environment and threats.

Intuitively we feel that autonomous systems are not only more complex, but autonomy introduces new problems that may require quite novel approach for design and development of such systems. This raises a question how much autonomous systems differ from non-autonomous ones and whether we can apply the same approach for safety assurance. It is not certain if the existing methods and techniques would be adequate and sufficient. Traditional safety analysis is based on system safety state model with transitions between states (e.g. from safe to hazardous state and then to accident state) where transitions can be initiated by external or internal events. For autonomous mobile systems (e.g. vehicles) its control system structure

Assuring safety of autonomous vehicles operating in an open environment requires reliable situation awareness, action planning and prediction of actions of other vehicles and objects. Factors that also have to be considered are certainty and completeness of available information and trust in information sources or other entities.

Approach:

The goal of the project was to explore the problem area, not to provide a solution. The work was focused on the analysis how safety analysis methods fit specific issues of autonomous systems and what strategies can be applied when constructing safety argumentation.

One approach to vehicle safety assurance is based on the assumption that hazardous sequences of events should be identified during hazard analysis and then some means of hazard avoidance and mitigation, like barriers, should be designed and implemented. Another approach is to design a system which is able to dynamically examine the risk associated with possible actions and then select the safest action to carry it out. Dynamic risk assessment requires maintaining the situation awareness and prediction of possible future situations. We analyse how these two approaches can be applied for autonomous vehicles and what strategies can be used for safety argumentation.

Schedule:

  • The project started in 2005

  • discontinued in 2008.

Results:

  • Situation awareness model as an element of autonomous control system architecture (high-level model and analysis)

  • Safety argumentation schemas for autonomous systems

Publications:

  1. A. Wardziński, Strategies for Autonomous Vehicles Safety Assurance, The 27th International Conference on Computer Safety, Reliability and Security SAFECOMP 2008, Newcastle upon Tyne (UK), 22-25 September 2008, Springer, Lecture Notes on Computer Science, vol. 5219, pp. 277-290
  2. Download PDF (431 kB)

  3. A. Wardziński, Dynamic Risk Assessment in Autonomous Vehicles Motion Planning, IEEE Conference on Information Technology, IT 2008, 19-21 May 2008, Gdańsk (Poland), pp. 127-130
  4. Download PDF (431 kB)

  5. A. Wardziński, The Role of Situation Awareness in Assuring Safety of Autonomous Vehicles, 25th International Conference on Computer Safety, Reliability and Security, SAFECOMP’2006, Gdańsk (Poland), September 2006, Springer, Lecture Notes on Computer Science, vol. 4166, pp. 205-218
  6. Download PDF (431 kB)