Search papers, labs, and topics across Lattice.
This paper investigates the role of tactile sensing in robotic disassembly tasks, comparing vision-only policies to those augmented with tactile RGB (TacRGB) and tactile force field (TacFF) data. They trained policies using a unified learning framework across simulated and real-world rigid and deformable disassembly tasks. The key finding is that TacFF-based policies consistently outperform vision-only and TacRGB policies, especially in contact-dependent and deformable settings, highlighting the importance of structured force-field representations.
Tactile force fields unlock robust robotic disassembly, outperforming vision-only and even naive RGB-tactile fusion, particularly in deformable and contact-rich scenarios.
Robotic disassembly involves contact-rich interactions in which successful manipulation depends not only on geometric alignment but also on force-dependent state transitions. While vision-based policies perform well in structured settings, their reliability often degrades in tight-tolerance, contact-dominated, or deformable scenarios. In this work, we systematically investigate the role of tactile sensing in robotic disassembly through both simulation and real-world experiments. We construct five rigid-body disassembly tasks in simulation with increasing geometric constraints and extraction difficulty. We further design five real-world tasks, including three rigid and two deformable scenarios, to evaluate contact-dependent manipulation. Within a unified learning framework, we compare three sensing configurations: Vision Only, Vision + tactile RGB (TacRGB), and Vision + tactile force field (TacFF). Across both simulation and real-world experiments, TacFF-based policies consistently achieve the highest success rates, with particularly notable gains in contact-dependent and deformable settings. Notably, naive fusion of TacRGB and TacFF underperforms either modality alone, indicating that simple concatenation can dilute task-relevant force information. Our results show that tactile sensing plays a critical, task-dependent role in robotic disassembly, with structured force-field representations being particularly effective in contact-dominated scenarios.
