Risk-aware graph search with dynamic edge cost discovery

Abstract

In this paper, we introduce a novel algorithm for incorporating uncertainty into lookahead planning. Our algorithm searches through connected graphs with uncertain edge costs represented by known probability distributions. As a robot moves through the graph, the true edge costs of adjacent edges are revealed to the planner prior to traversal. This locally revealed information allows the planner to improve performance by predicting the benefit of edge costs revealed in the future and updating the plan accordingly in an online manner. Our proposed algorithm, risk-aware graph search (RAGS), selects paths with high probability of yielding low costs based on the probability distributions of individual edge traversal costs. We analyze RAGS for its correctness and computational complexity and provide a bounding strategy to reduce its complexity. We then present results in an example search domain and report improved performance compared with traditional heuristic search techniques. Lastly, we implement the algorithm in both simulated missions and field trials using satellite imagery to demonstrate the benefits of risk-aware planning through uncertain terrain for low-flying unmanned aerial vehicles.

Multi-robot Coordination and Planning in Uncertain and Adversarial Environments

“…These represent extremes; recently, there is a trend of using more nuanced measures that better capture risk. In particular, these include mean-variance [7,73], chance-constraints or value-at-risk (VaR) [22 •], and conditional-value-at-risk (CVaR) [23••, 74 •, 75] (Fig. 2).…”

“…Based on these risk measures, researchers have developed risk-aware approaches in various robotics tasks such as graph search and motion planning [74,75,7,77], controls [8,78,25], task allocation and assignment [6,22,79,80], information collection [81,23], and machine learning [75,82]. Here, we focus on the risk-aware approaches for multi-robot systems.…”

“…As a result, any algorithm that coordinates the actions of the robots will have to operate on uncertain information. Coordination in the presence of uncertainty can be risky [6,7,9]. There is a risk of the actual performance of the robot team during execution time being significantly different from the expected performance during planning time.…”

Falco: Fast likelihood‐based collision avoidance with extension to human‐guided navigation

“…These methods involve probabilities to model the motion or state of the vehicle. Chung, Smith, Skeele, and Hollinger (2019) model the edge costs on a graph with uncertainties for graph search. Heiden, Hausman, Sukhatme, and Agha‐mohammadi (2017) use probabilities to model the traversability of map voxels.…”

A Foraging Strategy with Risk Response for Individual Robots in Adversarial Environments

“…Finally, this algorithm will traverse the entire hierarchical structure, and the route corresponding to the hierarchy with the highest utility will be selected as the inal foraging route. These algorithms are evaluated in simulations and compared to other existing benchmark algorithms for performance in robot foraging [18,75]. The results show that the proposed algorithms outperform the benchmark methods on scalability and solution quality.…”