A Model for Multi-Agent Autonomy That Uses Opinion Dynamics and Multi-Objective Behavior Optimization

Abstract

This paper reports a new hierarchical architecture for modeling autonomous multi-robot systems (MRSs): a nonlinear dynamical opinion process is used to model high-level group choice, and multi-objective behavior optimization is used to model individual decisions. Using previously reported theoretical results, we show it is possible to design the behavior of the MRS by the selection of a relatively small set of parameters. The resulting behavior - both collective actions and individual actions - can be understood intuitively. The approach is entirely decentralized and the communication cost scales by the number of group options, not agents. We demonstrated the effectiveness of this approach using a hypothetical `explore-exploit-migrate' scenario in a two hour field demonstration with eight unmanned surface vessels (USVs). The results from our preliminary field experiment show the collective behavior is robust even with time-varying network topology and agent dropouts.

Figures (7)

• Figure 1: Overall framework of group choice with individual decision (GCID). A networked system of robots is shown on the left, each sharing opinions with neighbors (blue). The group choice selects which local behaviors are active and the behavior optimization decides the reference input to the lower-level controller (yellow).
• Figure 2: Overview of Opinion Manager Engine which separately runs on each vehicle in a decentralized scheme. At each iteration, local opinion inputs, any available opinion of neighbors, and previous own opinion state are used to determine the next opinion state. The new state is communicated to any nearby vehicles.
• Figure 3: 8 Heron USVs used in a two-hour field deployment at night on the Charles River in Boston, MA
• Figure 4: State of the MRS and simulated environment approximately 50 minutes into a 2 hour mission on the Charles River near the MIT Sailing Pavilion. Simulated blooms appeared randomly within both zones X-ray and Yankee, and grew larger with time. A randomly generated storm periodically passed over the regions. At this time during the mission, 8 vehicles were searching and sampling in Zone X-ray. The communication range was artificially limited to 160 meters, and the restricted inter-vehicle communication is visualized in red. Both zones measured 300 meters by 350 meters.
• Figure 5: Monte Carlo simulation experiments of the bloom sampling scenario with increasing fleet size. The MRS using the GCID approach sampled a higher percentage of blooms on average (top), and also sampled blooms with a slightly higher average efficiency - greater samples per meter of travel (bottom). Solid lines are the mean for each type and the shaded areas indicate the min and max values of all trials.