Controller synthesis for autonomous systems interacting with human operators

Abstract

We propose an approach to synthesize control protocols for autonomous systems that account for uncertainties and imperfections in interactions with human operators. As an illustrative example, we consider a scenario involving road network surveillance by an unmanned aerial vehicle (UAV) that is controlled remotely by a human operator but also has a certain degree of autonomy. Depending on the type (i.e., probabilistic and/or nondeterministic) of knowledge about the uncertainties and imperfections in the operatorautonomy interactions, we use abstractions based on Markov decision processes and augment these models to stochastic two-player games. Our approach enables the synthesis of operator-dependent optimal mission plans for the UAV, highlighting the effects of operator characteristics (e.g., workload, proficiency, and fatigue) on UAV mission performance; it can also provide informative feedback (e.g., Pareto curves showing the trade-offs between multiple mission objectives), potentially assisting the operator in decision-making. ABSTRACTWe propose an approach to synthesize control protocols for autonomous systems that account for uncertainties and imperfections in interactions with human operators. As an illustrative example, we consider a scenario involving road network surveillance by an unmanned aerial vehicle (UAV) that is controlled remotely by a human operator but also has a certain degree of autonomy. Depending on the type (i.e., probabilistic and/or nondeterministic) of knowledge about the uncertainties and imperfections in the operatorautonomy interactions, we use abstractions based on Markov decision processes and augment these models to stochastic two-player games. Our approach enables the synthesis of operator-dependent optimal mission plans for the UAV, highlighting the e↵ects of operator characteristics (e.g., workload, proficiency, and fatigue) on UAV mission performance; it can also provide informative feedback (e.g., Pareto curves showing the trade-o↵s between multiple mission objectives), potentially assisting the operator in decision-making.

PRISM-games: verification and strategy synthesis for stochastic multi-player games with multiple objectives

“…Table 2 shows statistics for multi-objective strategy synthesis. The models are taken from four case studies: uav (human-in-the-loop UAV mission planning) [24], driving (autonomous urban driving) [16], power (aircraft power dis- tribution) [5] and temp (temperature control) [44]; the first three are discussed in the next section. The case studies do not all have parameters to scale the models, as in Table 1, but we show two variants of driving (using maps for two villages), and for power, we vary the switch delays d. The last two case studies are used for compositional (assume-guarantee) strategy synthesis.…”

“…Human-in-the-loop UAV mission planning [24] This case study concerns autonomous unmanned aerial vehicles (UAV) performing road network surveillance and reacting to inputs from a human operator. The UAV performs most of the piloting functions, such as selecting the waypoints and flying the route.…”

Risk and reliability modelling for multi-vehicle marine domains

“…Whilst focussing on plan generation, Feng et al [29] represent their multi-agent surveillance domain as both an MDP and a two-player stochastic game. The domain requires interaction between the two agents -a UAV flies between waypoints autonomously, while a human operator controls the capture of images at waypoints.…”

“…As in the work by Brito and Griffiths [16], Markov chains have been used as effective representation for simple AUV deployments, as well as more widely as representations for risk analysis [30]- [32]. However, as mission formats become more dynamic, with vehicle behaviour dependent on interaction between multiple platforms, the problem may be better represented as an MDP, as used by Feng et al [29]. An MDP would allow the choice of action to take at a given state to be modelled, e.g.…”

Analyzing Resilience Properties of Different Topologies of Collective Adaptive Systems

“…An SMG strategy resolves the nondeterministic choices in each state, selecting actions for a player based on the current state and a set of memory elements. 1 Probabilistic model checking of SMGs has been applied to a variety of analysis and synthesis problems [11], [12]. In the context of self-adaptive systems, we presented in previous work [13] an analysis technique based on model checking of SMGs to quantify the potential benefits of employing different types of algorithms for self-adaptation.…”