TERRA 框架让火星车在松软坡地上 3-5 次试跑内完成在线适应
Explore Where It Matters: Targeted Exploration for Efficient Rover Adaptation
机器人团队做了个 TERRA 框架，火星车到了松软坡地上不用大量试错，3-5 次真实试跑就能学会主动悬挂爬坡，仿真里成功率能到 90%。做机器人 sim-to-real 的可以看看。
arXiv 论文提出 TERRA（Targeted Exploration for Reward-Relevant Adaptation）在线适应框架，用于行星漫游车在可变形地形上的部署时适应。该方法用奖励加权的预测损失微调仿真训练的世界模型，并训练损失预测器引导 MPC 探索。仿真实验中，TERRA 在严重分布外预训练条件下 5 次试验内达到约 90% 的爬坡成功率。在真实 ERNEST 漫游车上，TERRA 在 3-5 次真实试验内恢复火星模拟土坡上的主动悬挂协调运动。
Explore Where It Matters: Targeted Exploration for Efficient Rover Adaptation
Future planetary rovers will traverse steep, deformable terrain where slip, sinkage, and terrain deformation can threaten mobility. Although simulation can capture some wheel-soil effects, unmodeled deformable-terrain phenomena create sim-to-real gaps requiring deployment-time adaptation. Yet, rover operations leave little room for extensive trial-and-error. We introduce TERRA (Targeted Exploration for Reward-Relevant Adaptation), an online adaptation framework that efficiently seeks deployment data exposing task-relevant model errors and corrects them. TERRA adapts a simulation-trained world model with a reward-weighted prediction loss and trains a loss predictor to estimate where future performance-relevant errors are likely to occur, using this signal to guide exploration during MPC. In simulation, TERRA reaches approximately 90% slope-climbing success within five trials under severe out-of-distribution pretraining conditions, outperforming exploration baselines in reward, completion time, and energy to first success. On the ERNEST rover, TERRA recovers coordinated active-suspension locomotion on a deformable Mars-simulant slope within three to five real-world trials. Project website: https://terra-explore.github.io