Machine Learning Based Path Planning for Improved Rover Navigation

Abstract

Enhanced AutoNav (ENav), the baseline surface navigation software for NASA's Perseverance rover, sorts a list of candidate paths for the rover to traverse, then uses the Approximate Clearance Evaluation (ACE) algorithm to evaluate whether the most highly ranked paths are safe. ACE is crucial for maintaining the safety of the rover, but is computationally expensive. If the most promising candidates in the list of paths are all found to be infeasible, ENav must continue to search the list and run time-consuming ACE evaluations until a feasible path is found. In this paper, we present two heuristics that, given a terrain heightmap around the rover, produce cost estimates that more effectively rank the candidate paths before ACE evaluation. The first heuristic uses Sobel operators and convolution to incorporate the cost of traversing high-gradient terrain. The second heuristic uses a machine learning (ML) model to predict areas that will be deemed untraversable by ACE. We used physics simulations to collect training data for the ML model and to run Monte Carlo trials to quantify navigation performance across a variety of terrains with various slopes and rock distributions. Compared to ENav's baseline performance, integrating the heuristics can lead to a significant reduction in ACE evaluations and average computation time per planning cycle, increase path efficiency, and maintain or improve the rate of successful traverses. This strategy of targeting specific bottlenecks with ML while maintaining the original ACE safety checks provides an example of how ML can be infused into planetary science missions and other safety-critical software.

MLNav: Learning to Safely Navigate on Martian Terrains

“…In a nutshell, MLNav is a search-based path planner that uses learned heuristics, where the safety of the chosen path is guaranteed by running a model-based collision checker. The use of learned heuristics within search-based robotic planning has previously been studied [7]- [11], including showing favorable comparisons versus hand-craft heuristics for rover path planning [12]. Our main contribution is proposing a general system design principle for effectively integrating ML-based approaches into existing navigation pipelines of safety-critical robotic systems, as well as a concrete instantiation for Mars rover navigation.…”

Semi-Supervised Learning for Mars Imagery Classification and Segmentation

“…The Opportunistic Rover Science (OASIS) framework uses machine learning algorithms to identify terrain features [8,20], dust devils, and clouds [7]. For rover navigation, Abcouwer et al [1] presented two heuristics to rank candidate paths, where a machine learning model is applied to predict untraversable areas. For ex-situ methods, machine learning can help scientists analyze data and notify noteworthy findings.…”

Machine Vision based Sample-Tube Localization for Mars Sample Return

“…As a result they are unable to handle texture-less objects like sample-tubes. More recently, with the advent of deep learning, CNNs by comparison, have made significant progress in object classification [25], detection [26], [27], semantic segmentation [28] and instance segmentation [29], including their application to Mars rover autonomy [30], [31].…”