The PhD work of Alexis was presented at Eurohaptics 2026. It contributes to haptic guidance in teleoperation with guidelines to properly define and select a haptic guidance model and to evaluate the human-guidance interaction.

Abstract

Haptic guidance in teleoperation enhances operator performance through force feedback. This paper presents guidelines to select the most appropriate model considering the task, the environment and the operator. We define a unified \added{formulation} expressing most common models (spring-damper, potential field, and guiding tube) as variations of a stiffness-damping system with model-specific guiding functions. We conducted a user study comparing the three classical models across six scenarios with varying environmental conditions in a vertical farming task. Results show no universally superior model: spring-damper excels in cluttered environments, potential field in free spaces (but it shows risks near obstacles), and guiding tube offers a balanced compromise. We propose novel objective metrics to evaluate the interaction, and show that guiding force magnitude correlates with comfort and trust scores. These findings provide practical model selection guidelines through environmental characteristics and real-time evaluation metrics.

The PhD work of Elio was presented at ICRA 2026. It contributes to blending in shared control with a Model Predictive Control approach that enforces safety and improves the assistance quality.

Abstract

Shared control methods distribute control between human operators and robots in demanding tasks, enabling collaboration that leverages their respective strengths and expertise. Sharing the task typically involves blending algorithms that combine human control inputs to (pre)planed assistance trajectories. Conventional blending techniques, such as Linear Blending, compute a combined output but neither guarantee feasibility of this shared motion, nor ensure compliance with safety or task-related constraints. This paper proposes to tackle feasibility and safety by formulating the blending strategy as the solution of a constrained optimal control problem, that enforces environment limits, task requirements, and physical capabilities. A Model Predictive Control approach is used to solve the optimization problem and anticipate constraints by predicting the robot motion over a receding time horizon. We evaluate this approach in simulated and real-world pick-and-place teleoperation experiments. The experimental study compares the Model Predictive Control approach to Linear Blending and full Teleoperation. The results show that the new framework offers significant improvements, as it provides a safer, more accurate, and repeatable response.

Congratulations to Pierre for this contribution advancing fast and clinically applicable upper-limb musculoskeletal modeling.

Abstract

This paper addresses the challenge of estimating personalized muscle forces through musculoskeletal modeling, which is valuable for assessing patient status and monitoring clinical progress. Upper-limb applications have been limited due to system complexity and the long computation times of existing calibration methods. We propose a fast (<5 min) calibration method for upper-limb models, calibrating maximal isometric force and optimal muscle length for 38 muscles across 10 degrees of freedom by matching muscle-generated moments with dynamically consistent joint moments. The method leverages experimental data including bony landmark trajectories from markerless motion capture, external forces, and electromyography (EMG). During hand-cycling, the calibrated model reduced EMG tracking error compared to the uncalibrated model (5.58±0.92% vs. 6.30±1.28%), and reliance on non-physiological residual moments was also lowered (12.68 vs. 23.61% of peak moment). This approach provides a fast and reliable framework for upper-limb musculoskeletal calibration, facilitating more accurate and clinically applicable muscle force estimation.

This work contributes a practical and open tool for harmonising scapular kinematic data across studies, addressing a long-standing interoperability challenge in shoulder biomechanics.

Abstract

Defining bone-embedded local coordinate systems (LCSs) is fundamental in shoulder biomechanics, yet multiple scapular LCSs exist, hindering data comparison. Although the International Society of Biomechanics (ISB) has published recommendations, alternative definitions based on distinct anatomical landmarks remain widely used. To address these inconsistencies, we extend a previously proposed average rotation matrix approach to 11 scapular LCSs reported in the literature, including both scapula- and glenoid-based systems. Using statistical shape models derived from 80 participants (asymptomatic and pathological shoulders), 1000 scapulae were generated to quantify geometric transformations between LCSs. The application of the average rotation matrices substantially reduced the maximal discrepancies between LCSs, from 21.9° to 5.2°. Scapula-based systems exhibited lower discrepancies than glenoid-based ones, reflecting greater morphological variability in the glenoid region. These findings confirm that average rotation matrices provide a reliable means of harmonising scapular kinematic data across studies, promoting interoperability in shoulder biomechanics research.

This paper is one of the many positive outcomes of our collaboration with the Aerospline company through the Plan de relance program. Congratulations to Guillaume for his dedication on this work.

Abstract

This paper addresses the challenging problem of Semi-Constrained End-Effector Path Planning for robotic manipulators. This problem arises when complex specifications restrict the end-effector’s motion during the execution of industrial tasks. Traditional path planning algorithms often struggle with such problems due to the difficulty of exploring the robot’s valid configuration space, or constrained manifold, under these conditions. In this work, we propose a novel sampling-based approach that efficiently navigates the constrained manifold by exploring an alternative space representing the end-effector’s degrees of freedom, such as process-related tolerances, throughout the task. This method retains the simplicity of sampling-based techniques. Building on this approach, we introduce the F-RRT algorithm, an adaptation of the renowned RRT planner [1]. F-RRT demonstrates enhanced speed and robustness compared to existing solutions, particularly in complex and cluttered environments.

Our paper based on the PhD work of Antun Skuric and Pycapacity on generating online optimal robot motions that best exploit the robot capabilities has been accepted for publication in the IEEE Transactions on Robotics.

Abstract

Conforming to safety standards often limits collaborative robots’ performance and size, restricting their applications despite their capabilities. Planning their motions in human environments involves a trade-off between optimal trajectory planning and quick adaptation to dynamic, unstructured spaces. Traditional trajectory planning methods either use simplified robot models and sacrifice robot’s abilities for computational efficiency, or exploit robots’ abilities fully but have high computational complexity and rely on substantial pre-computation. This paper introduces an approach for trajectory planning that exploits robot’s full motion abilities while planning on-the-fly. In each step of the trajectory execution, it evaluates robot’s movement ability using polytope algebra and calculates a time-optimal Trapezoidal Acceleration Profile (TAP) on the remaining trajectory. The method is shown to be near time-optimal (around 5% slower trajectories) by benchmarking it against the state-of-the-art time-optimal method TOPP-RA. The method allows reaching higher velocities (able to plan up to 100% of the robot’s kinematic limits) while at the same time lowering the tracking error (under 4mm) than traditional Cartesian Space planning methods. A mock-up experiment demonstrates its efficiency in collaborative waste sorting using a Franka Emika Panda robot.

Rémi Lafitte presented our works about modeling of human visuo-haptic perception at the 23rd International Multisensory Research Forum in Durham (GB).

Abstract

Reliable visual and haptic feedback are known to improve teleoperation tasks. Better understanding human visuo-haptic perception could help to develop feedback strategies that better inform the human operator during human-robot interaction. The project is conducted in collaboration with the CeRCA-CNRS at the University of Poitiers.

Psychophysic experiments have suggested that the brain combines visual and haptic estimates of environmental properties (e.g., object size) in a statistically optimal fashion. Whether this sensory integration still holds in a more challenging environment, such as for teleoperation, remains unknown.

Paper Abstract

In hazardous or inaccessible environments, teleoperation enables remote task execution. However, its effectiveness is hindered by reduced performance and an increased operator workload, primarily due to the lack of direct sensory feedback. To address this issue, haptic guidance, that provides a guiding force feedback through the haptic device, can enhance remote operations by directing the operator towards task goals. This method relies on both the environment and the operator, and its effectiveness is affected by their dynamics. Consequently, developing an adaptive guidance that responds to changing situations is essential for practical use.

Teleoperation is a method often used to carry out tasks in dangerous, inaccessible or sensitive environments where human’s expertise and ability to adapt cannot be replaced by an autonomous robot. However, teleoperation also involves a split between the operator and the workspace. To counter this, haptic interfaces are used, transmitting the physical interactions of the robot with its environment via force feedback. The facility to generate forces, on the operator side, can also be exploited to help perform a task by the use of guiding forces.

Paper Abstract

Robots require the ability to autonomously and continuously react to unexpected online changes in the task definition and in the environment, especially those cohabited with humans. To react to these changes, the task, from the current state up to the finish, must instantly be reconsidered. This implies a prohibitive re-computation cost.

This paper proposes a modular control architecture based on Model Predictive Control, that offers a good compromise between optimally achieving the task and the required computation time, by only reconsidering the near future. This framework offers a generic way to formulate task-related objectives and constraints that dissociates the planning from the execution, which depends mainly on the robot dynamics.