A Map That Isn't Afraid to Admit: 'I Don't Know' express
A group of robot explorers is like a team of cooks tasting the same dish. Each samples their own portion and shares an impression: clear, occupied, or uncertain. But tastes are subjective — one might miss a chair leg, another might only lick the edge. Instead of averaging the discord, the robots agree: where confidence is low, honestly write 'don't know.' To do this, they exchange an e-value — a measure of their own uncertainty (how doubtful their report is). Pooling all these 'doubt numbers,' they get an overall picture where reliability is mathematically guaranteed — say, 95%. The more 'cooks' and the more often they check, the fewer dark spots remain in the recipe. The method doesn't depend on who talks to whom or how bad the sensors are: a robot is allowed to be wrong if it honestly admits it.
Original abstract
as written by the authors arXiv:2607.14906
Accurate and reliable environmental mapping is a fundamental requirement for multi-robot autonomy. While continuous mapping techniques like Gaussian Process Occupancy Mapping (GPOM) provide rich spatial correlation and uncertainty estimates, they lack formal, finite-sample guarantees on their predictive reliability. Conformal prediction can equip each robot's local map with a distribution-free coverage guarantee, but this local guarantee degrades in practice: temporal correlation along a robot's trajectory breaks the exchangeability on which conformal calibration relies, and each robot observes only a spatially limited, non-uniform portion of the environment. Taking these degraded per-agent guarantees as given, we develop a distributed fusion algorithm that recovers the desired coverage across the team. Robots exchange only lightweight scalar e-values with their neighbors, and a receiver fuses them using a per-neighborhood miscoverage budget and an uncertainty-attenuated fusion operator. We prove that the fused set-valued map recovers the target user-specified coverage level regardless of the communication graph topology or the underlying sensor noise distribution. However, a drawback is that wherever the fused evidence is insufficient, the map declines to commit and returns both labels (free and occupied), leaving a significant fraction of the domain unclassified rather than thresholded into a single decision. Simulated multi-agent mapping experiments demonstrate that the fused predictor reliably meets its theoretical coverage bounds, and illustrate that denser communication topologies significantly enhance map efficiency by shrinking this unclassified fraction.
The most accurate map is the one that doesn't shy away from gaps: a robot that doubts its readings is more useful than an overconfident know-it-all.