Long-Term Autonomy in Forest Environment using Self-Corrective SLAM

TL;DR

This paper presents a novel SLAM approach for long-term forest environment autonomy, combining lightweight edge computation with cloud-based corrections to improve map accuracy over extended missions.

Contribution

It introduces a self-corrective SLAM method that interpolates transformations and incorporates iterative ICP corrections, enhancing long-term mapping accuracy in forest environments.

Findings

Adding 4% more ICP matches reduces RMSE to 0.15 m

The method effectively maintains map quality over 180 m odometric distance

Cloud-based corrections improve long-term SLAM accuracy

Abstract

Vehicles with prolonged autonomous missions have to maintain environment awareness by simultaneous localization and mapping (SLAM). Closed loop correction is substituted by interpolation in rigid body transformation space in order to systematically reduce the accumulated error over different scales. The computation is divided to an edge computed lightweight SLAM and iterative corrections in the cloud environment. Tree locations in the forest environment are sent via a potentially limited communication bandwidths. Data from a real forest site is used in the verification of the proposed algorithm. The algorithm adds new iterative closest point (ICP) cases to the initial SLAM and measures the resulting map quality by the mean of the root mean squared error (RMSE) of individual tree clusters. Adding 4% more match cases yields the mean RMSE 0.15 m on a large site with 180 m odometric distance.