Online Risk Management for milliAmpere 2

Background & motivation

  • A major part of a human navigator’s role involves continuously evaluating the risk level and make risk-aware decisions to ensure safe operation.
  • When replacing a human navigator with an autonomy system, the autonomy itself must perform the risk analysis and make decisions accordingly. We refer to this as Online Risk Management (ORM). 
  • ORM can enable monitoring of internal systems and the operative environment, calculation of risk levels for a set of hazards, provisioning of early warnings and tactical decision-making.

Problem description

  • Utne et al. (2020) har proposed a method for developing an online risk model using Systems Theoretic Process Analysis (STPA) and Bayesian Belief Networks (BBNs). The student should develop an ORM system using this methodology for the milliAmpere 2 autonomous passenger ferry.

Work proposal

  • The student will be given access to risk analyses for milliAmpere 2 using STPA and HAZID. This should be used to create an online risk model in the form of a BBN, as proposed by Utne et al. (2020). 
  • The student should investigate how this online risk model can be applied for online risk management for milliAmpere 2. Possible uses include:
    • Online decision making
      • Cross/stay in dock
      • Enter/Leave minimum risk condition (MRC)
      • Select MRC
    • Adaptation of control based on risk level
      • Crossing speed
      • Safety margins in collision avoidance
    • Give alarms/early warning to shore control centre
  • The proposed ORM should be implemented and interfaced to the milliAmpere 2 simulator developed by NTNU and Zeabuzand a simulation study to evaluate it should be conducted. 

Figure 1: This figure illustrates a method for developing an online risk model using Systems Theoretic Process Analysis (STPA) and Bayesian Belief Networks (BBNs)