Explicit compliance and safety on torque controlled robots for physical interaction
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
This paper introduces a new paradigm of compliant control for torque-controlled robots through the notion of explicit compliance, which enables selective and tunable compliance for each task or joint in a control hierarchy. We reformulate the motion generation quadratic program (QP) to incorporate this explicit compliance model, allowing the robot to adaptively respond to external forces while preserving task priorities. Our formulation also integrates safety and feasibility constraints—such as torque, velocity, and self-collision limits—at the highest level of the control hierarchy. To improve robustness near constraint boundaries, we propose a second-order velocity damper expressed in acceleration, which ensures stable constraint enforcement without dependency on the control loop frequency. In addition, we enhance external force estimation through a lag-free sensor fusion strategy that combines high-frequency force/torque sensor measurements with low-frequency residual-based estimates. This complementary filter achieves accurate external torque estimation across contact scenarios, reducing the RMS estimation error by about 40% from 11.168 N (residual only) to 6.949 N. The proposed framework is deployed on a Kinova Gen3 robot and validated through experiments with various compliance configurations. Using the compliance parameter Γ, we demonstrate three distinct behaviors: full stiffness, null-space compliance, and full-body compliance. Our results show that the proposed approach preserves safety under contact while offering precise task execution and flexible compliance, enabling safe and adaptable physical interaction in dynamic environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it