When to Ask for Help: Uncertainty-Triggered Aerial Refinement for Off-Road Navigation
When to Ask for Help: Uncertainty-Triggered Aerial Refinement for Off-Road Navigation
This research introduces a method for heterogeneous air-ground teams where aerial sensing addresses ground-vehicle sensor limitations. The approach evaluates how uncertain map information affects the UGV's current MPPI rollouts and requests aerial measurements only when resolving uncertainty improves predicted navigation. Across 180 simulation trials on three off-road routes, the method achieves 98.9% goal attainment while reducing mean UAV flight time by 87.7% relative to continuous greedy aerial mapping. Field experiments spanning approximately 2 km of UGV operation demonstrate successful navigation recovery.
The paper proposes a planner-conditioned method for aerial assistance to ground vehicles that triggers UAV measurements only when resolving map uncertainty improves predicted navigation, achieving 98.9% goal attainment in 180 simulation trials and demonstrating field recovery over approximately 2 km of UGV operation.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S4
Reported numbers
UGV navigation time saved per UAV second
2.94
View original evidence
edicted ground navigation. Across 180 simulation trials on three off-road routes, the method achieves 98.9% goal attainment while reducing mean UAV flight time by 87.7% relative to continuous greedy aerial mapping. Aggregated across the routes, each second of estimated UAV flight time saves 2.94 s of UGV navigation time. Field experiments spanning approximately 2 km of UGV operation demonstrate uncertainty resolution
Open source S4simulation trials count
180
Reported trials
View original evidence
edicted ground navigation. Across 180 simulation trials on three off-road routes, the method achieves 98.9% goal attainment while reducing mean UAV flight time by 87.7% relative to continuous greedy aerial mapping. Aggregated across the routes, each second of estimated UAV flight time saves 2.94 s of UGV navigation time. Field experiments spanning approximately 2 km of UGV operation demonstrate uncertainty resolution
Open source S4UGV operation distance
≈2
View original evidence
edicted ground navigation. Across 180 simulation trials on three off-road routes, the method achieves 98.9% goal attainment while reducing mean UAV flight time by 87.7% relative to continuous greedy aerial mapping. Aggregated across the routes, each second of estimated UAV flight time saves 2.94 s of UGV navigation time. Field experiments spanning approximately 2 km of UGV operation demonstrate uncertainty resolution
Open source S4
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
edicted ground navigation. Across 180 simulation trials on three off-road routes, the method achieves 98.9% goal attainment while reducing mean UAV flight time by 87.7% relative to continuous greedy aerial mapping. Aggregated across the routes, each second of estimated UAV flight time saves 2.94 s of UGV navigation time. Field experiments spanning approximately 2 km of UGV operation demonstrate uncertainty resolution
Open source S4
When to Ask for Help: Uncertainty-Triggered Aerial Refinement for Off-Road Navigation
Open source S1
arXiv:2611.06702v1 Announce Type: new Abstract: Heterogeneous air-ground teams can use aerial sensing to address the limited range and viewpoint of ground-vehicle sensors, but deciding when the UGV needs assistance and what information the UAV should collect remains a challenge. This paper presents a planner-conditioned air-ground assistance method that evaluates how uncertain map information affects the UGV's curren
Open source S2
t MPPI rollouts. When navigation remains degraded, we score the same rollouts under resolved map hypotheses, select the map layer with the largest useful improvement, request the corresponding aerial measurement, fuse the returned information, and replan. Unlike approaches driven by fixed support roles, task structure, or uncertainty alone, our method conditions assistance on whether resolving uncertainty improves pr
Open source S3
Source：arXiv Robotics — research abstracts · arxiv.org