SenseSwarm: Swarm Exploration via Neural Structure Estimation with Shared World-Anticipation for Rapid Mapping
SenseSwarm: Swarm Exploration via Neural Structure Estimation with Shared World-Anticipation for Rapid Mapping
SenseSwarm is a research paper proposing a decentralized swarm exploration method that uses a lightweight learned map predictor to enhance mapping efficiency. The approach integrates neural structure estimation with shared world-anticipation, allowing a swarm of robots to rapidly map unknown environments. The method is evaluated on a benchmark of real building floors and simulations, showing improved performance in map reconstruction compared to non-predictive baselines.
SenseSwarm introduces a novel approach to swarm exploration by integrating neural structure estimation with shared world-anticipation, enabling faster and more accurate mapping of unknown environments. The method demonstrates significant improvements in map reconstruction efficiency compared to non-predictive baselines.
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.
Reported numbers
square meters
≤24,000
View original evidence
ucture back into a stigmergic guidance field. Its 709K-parameter predictor is fine-tuned on the swarm's own multi-front observations, which we show is necessary: a single-robot predictor collapses on such views. We build a benchmark of twelve real building floors from HouseExpo, HKUST, and KTH, up to 24,000 square meters, and run 1,620 simulations across nine coordination methods, swarm sizes 32 to 512, and three see
Open source S3square meters
≈670
View original evidence
ds. On physical coverage the swarm is near parity with a greedy frontier planner, as expected. On map reconstruction, every non-predictive baseline needs 1.3 to 1.9 times longer to reach an 80 percent known map, and SenseSwarm holds an equal-time area lead of about 670 square meters, up to 1,359. Folding the prediction into the shared field matters: the same predictor applied per robot matches our speed but suffers a
Open source S4square meters
≤1,359
View original evidence
ds. On physical coverage the swarm is near parity with a greedy frontier planner, as expected. On map reconstruction, every non-predictive baseline needs 1.3 to 1.9 times longer to reach an 80 percent known map, and SenseSwarm holds an equal-time area lead of about 670 square meters, up to 1,359. Folding the prediction into the shared field matters: the same predictor applied per robot matches our speed but suffers a
Open source S4Robot count
512 robots
View original evidence
6.9 times higher inter-robot contact rate at 512 robots. The prediction is computed once per shared map, so its cost stayed flat from 32 to 512 robots.
Open source S5Robot count
32–512 robots
View original evidence
6.9 times higher inter-robot contact rate at 512 robots. The prediction is computed once per shared map, so its cost stayed flat from 32 to 512 robots.
Open source S5
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
ucture back into a stigmergic guidance field. Its 709K-parameter predictor is fine-tuned on the swarm's own multi-front observations, which we show is necessary: a single-robot predictor collapses on such views. We build a benchmark of twelve real building floors from HouseExpo, HKUST, and KTH, up to 24,000 square meters, and run 1,620 simulations across nine coordination methods, swarm sizes 32 to 512, and three see
Open source S3
ds. On physical coverage the swarm is near parity with a greedy frontier planner, as expected. On map reconstruction, every non-predictive baseline needs 1.3 to 1.9 times longer to reach an 80 percent known map, and SenseSwarm holds an equal-time area lead of about 670 square meters, up to 1,359. Folding the prediction into the shared field matters: the same predictor applied per robot matches our speed but suffers a
Open source S4
6.9 times higher inter-robot contact rate at 512 robots. The prediction is computed once per shared map, so its cost stayed flat from 32 to 512 robots.
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org