Presentation for the IEEE International Conference on Robotics and Automation (ICRA) 2022

Julian Nubert, Shehryar Khattak and Marco Hutter

Paper: https://arxiv.org/pdf/2203.01389.pdf
Code: https://github.com/leggedrobotics/GMFCL

Abstract:
Enabling autonomous operation of largescale construction machines, such as excavators, can bring key benefits for human safety and operational opportunities for applications in dangerous and hazardous environments. To facilitate robot autonomy, robust and accurate stateestimation remains a core component to enable these machines for operation in a diverse set of complex environments. In this work, a method for multimodal sensor fusion for robot stateestimation and localization is presented, enabling operation of construction robots in realworld scenarios. The proposed approach presents a graphbased predictionupdate loop that combines the benefits of filtering and smoothing in order to provide consistent state estimates at high update rate, while maintaining accurate global localization for largescale earthmoving excavators. Furthermore, the proposed approach enables a flexible integration of asynchronous sensor measurements and provides consistent pose estimates even during phases of sensor dropout. For this purpose, a dualgraph design for switching between two distinct optimization problems is proposed, directly addressing temporary failure and the subsequent return of global position estimates. The proposed approach is implemented onboard two Menzi Muck walking excavators and validated during realworld tests conducted in representative operational environments.