Abstract
This paper discusses experiences, risks and mitigation in autonomous transport systems, to improve learning and risk-based governance. This is based on literature reviews, experiences of autonomous transport systems and involvement in the regulatory process. The maturity varies between rail, road, aviation and shipping. Issues that may be transferred and adapted between modes are selected. The three research questions are: to describe major risks introduced by autonomous transport systems; how to mitigate the main risks through design and regulation; and suggest a way forward. The risk is dependent on operational domain, surrounding infrastructure, meaningful human control (including control centrals with human based interfaces), sensor ability, reliable propulsion and technology that are secure. Meaningful human control and system certification is a challenge. There is a need to gather data from incidents. In Norway autonomous shipping and road transportation are prioritized. Autonomy in rail and aviation, has not been prioritized in the same manner.
Highlights
This paper discusses experiences of autonomous transport systems, to establish a framework for riskbased design and governance
In the following, we have documented experiences from eco-systems/software systems in autonomy; and a summary of key issues of autonomous systems used in road; air; sea and railroad
There is a need to reduce the complexity of operations (ODD),; build supporting infrastructure to ensure that the ship is aware of position and surroundings; perform a risk based design to mitigate main challenges; ensure third party verification; establish taxonomies for incident reporting and gather data of incidents and accidents to be shared in an open manner; design control centres based on the science of Human Factors in order to ensure meaningful human control
Summary
This paper discusses experiences of autonomous transport systems, to establish a framework for riskbased design and governance. A key definition in [12] is scope, i.e. Operational Design Domain (ODD) for the autonomous vehicle describing the area of operation (i.e. selected specific road or area), type of road/infrastructure needed, environmental constraints (weather), other allowed traffic in the same area, speed limits, allowed time of operations. This is a key definition of interest in all modes: sea/ air/ road/ rail. C4: Tire strike only with little risk element (e.g., clipping a curb during a tight turn), considered to be of such minimal risk that most drivers would not consider these incidents to be crashes
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have