Organizations: State Key Laboratory of Mechanical System and Vibration, Institute of Robotics, School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai, 200240, China
Collaborative robots are increasingly deployed in industrial scenarios characterized by frequent product changeovers. As an intuitive programming method, kinesthetic teaching facilitates rapid robot deployment. However, users may overlook the configuration of the robot during kinesthetic teaching, leading to degradation in operational performance. Operational performance refers to the capability of the robot to generate motion and can be quantified by the Minimum Singular Value of the Jacobian matrix. To address this issue, this paper proposes an online dynamic force guidance method that integrates performance constraint and optimization mechanisms. Specifically, variable admittance control maintains the operational performance of the robot above a predefined threshold, while a virtual force actively guides the user to drag the robot towards configurations with improved performance. Experiments are conducted on a 6-DOF collaborative robot, comparing three typical paths in the task space. To evaluate the quality of the taught trajectories, trajectory playback experiments are conducted to analyze the relationship between the operational performance of the robot and the work efficiency. The results demonstrate that the proposed method effectively enhances the operational performance of the robot and consequently improves the work efficiency, holding significant value for reducing production takt time in industrial deployment.
Figures & tables
Figure 1: Paths crossing and bypassing the singularity. The red dashed line indicates the shoulder singularity, which is a cylindrical surface centered at the base coordinate origin.
Figure 2: Block diagram of the control structure. The performance constraint module dynamically adjusts the controller parameters based on the robot state. The performance optimization module generates a virtual force, which is combined with the operational force and sent to the adjusted admittance controller. The controller outputs the desired end-effector velocity, which is converted to joint velocity via the inverse Jacobian matrix and sent to the robot.
Figure 3: Influence of control parameters on dynamic force. (a) Admittance parameters. (b) Virtual force. The control parameters can be reasonably selected according to task requirements to adjust the magnitude and effective range of the dynamic force.
Figure 4: The mechanism of the dynamic force guidance. (a) Kinesthetic teaching with dynamic force guidance. (b) Control principle.
Figure 5: Experimental platform.
Performance constraint
Performance optimization
Parameter
Value
Parameter
Value
ζH
100
fmax(N)
15
ζL
15
k
24
σ0
96
lH
1
lL
0.5
Admittance parameters
Table 1: Control parameters used in the experiment
Figure 6: Experimental setup and performance indices. (a) Guide rails for three typical paths. (b) Taught trajectories and corresponding performance indices.
Figure 7: Real-world kinesthetic teaching demonstrations. (a) DT strategy on Path 2. (b) DFGT strategy on Path 2. (c) DT strategy on Path 3. (d) DFGT strategy on Path 3. The situation for Path 1 is similar to that of Path 2. The DT strategy reaches the singularity under both Path 1 and Path 3, leading to the termination of the kinesthetic teaching process.
Figure 8: Joint velocity and virtual force under different control strategies. (a) Direct Teaching. (b) Variable Admittance Teaching. (c) Dynamic Force Guided Teaching. Each row corresponds to one predefined path.
Figure 9: Statistical results of evaluation metrics for different control strategies on Path 2. (a) Trajectory length. (b) Minimum Singular Value. (c) Actual execution time. Asterisks indicate statistical significance: ∗∗p<0.001 .
Strategy
Length (mm)
MSV
Tex (s)
Fhmax (N)
DT
420.22
23.92
7.66
4.76
VAT
435.80
70.83
7.00
8.83
DFGT
444.62
77.15
6.87
6.96
Table 2: Evaluation Metrics under Different Control Strategies on Path 2
Figure 10: End-effector velocity and MSV over time during the playback process. (a) Direct Teaching. (b) Dynamic Force Guided Teaching.
Real-world reinforcement learning (RL) systems still struggle with the demands of contact-rich industrial manipulation, including micrometer-level precision, success rates above 99%, and human-level cycle times. Although off-policy algorithms can improve performance by leveraging demonstrations and interventions, a key bottleneck is the lack of an intuitive interface for collecting such guidance while complying with constraints of the physical system and the policy. We propose PAKT, a framework for kinesthetic teaching in RL. As opposed to teleoperation approaches, PAKT relies on kinesthetic guidance, which is widely used in industry. However, a critical weakness of kinesthetic guidance is the possibility for the operator to move the robot along trajectories (e.g., velocities, accelerations, jerk) that the robot and/or policy cannot physically reproduce. Using PAKT, operators guide the robot through admittance control, which maps human-applied forces to motion. The downstream reference generator applies the same kinematic limits used during policy execution, keeping the collected trajectories within these limits. To support this teaching interface with an appropriate execution layer, PAKT adds a high-performance control stack that maps low-frequency RL actions to high-frequency torque commands. It consists of a reference generator and subsequent impedance controller, where the reference generator preserves the tracking performance of the impedance controller while improving contact handling and producing smoother policy actions. Across the reported runs on four insertion and industrial assembly benchmarks, including a data center compute tray, the end-to-end system reduces cycle time by 23%-48% and cumulative intervention count by 62%-86% relative to the HIL-SERL baseline. Project website: https://pakt-website.github.io/pakt-website}{https://pakt-website.github.io/pakt-website
Instructing robots from demonstrations can be done through different teaching modalities, each with different usability and performance trade-offs. This paper compares kinesthetic guidance, joystick teleoperation, and hand gestures in a user study with eight participants. We evaluate replay success, modified NASA-TLX workload, and common teaching errors across three manipulation tasks. Kinesthetic guidance produced the shortest demonstrations, lowest workload, and highest success on the more orientation-sensitive and contact-rich tasks. Joystick teleoperation performed best on simple peg picking. Hand-gesture teaching, although less reliable overall, performed better than expected and in some cases achieved results comparable to kinesthetic guidance.
Petr Vanc, Jan Kristof Behrens, Václav Hlaváč +1
Czech Institute of Informatics, Robotics and Cybernetics (CIIRC CTU) Jugoslávských partyzánů 1580/3, 160 00 Dejvice
This work presents a human-centered collaborative framework that integrates Preference-Based Optimization (PBO) and Dynamic Movement Primitives (DMPs) to optimize robot-assisted tasks such as painting. The system allows the operator to perform the process while the robot adapts its behavior in real-time, dynamically adjusting the orientation of the piece in order to match the orientation of the operator's hand. The PBO framework leverages the GLISp algorithm to iteratively refine control parameters such as execution time, robot responsiveness, and rotation amplification through human feedback. Moreover, DMPs have been modified to enhance the reactive behavior of the robot and its adaptability to ergonomic requirements. The method was validated with a heterogeneous group of participants executing \rev{painting tasks}. The results show that our strategy effectively reduces operator effort while optimizing process outcomes.