SPARK-Remote: A Cost-Effective System for Remote Bimanual Robot Teleoperation

TL;DR

SPARK-Remote addresses the challenge of achieving precise, bimanual robot teleoperation at low cost by introducing SPARK, a 1:2 scaled, open-source platform, and SPARK-Remote, which adds force feedback and a force controller for remote operation. The approach balances affordability with performance by using a force glove and an optimization-based controller to mitigate high-torque events, evaluated across five bimanual tasks and multiple remote variants. Results show that combining force feedback and force control (SPARK-Remote-FGC) yields the best remote performance, closely approaching in-person SPARK in several tasks while keeping costs dramatically lower than traditional bilateral systems. The work demonstrates that affordable, multi-modal feedback can substantially reduce the remote-performance gap and highlights directions for future improvements, including multi-view visualization and higher-fidelity tactile feedback, enabled by open-source hardware designs.

Abstract

Robot teleoperation enables human control over robotic systems in environments where full autonomy is challenging. Recent advancements in low-cost teleoperation devices and VR/AR technologies have expanded accessibility, particularly for bimanual robot manipulators. However, transitioning from in-person to remote teleoperation presents challenges in task performance. We introduce SPARK, a kinematically scaled, low-cost teleoperation system for operating bimanual robots. Its effectiveness is compared to existing technologies like the 3D SpaceMouse and VR/AR controllers. We further extend SPARK to SPARK-Remote, integrating sensor-based force feedback using haptic gloves and a force controller for remote teleoperation. We evaluate SPARK and SPARK-Remote variants on 5 bimanual manipulation tasks which feature operational properties - positional precision, rotational precision, large movements in the workspace, and bimanual collaboration - to test the effective teleoperation modes. Our findings offer insights into improving low-cost teleoperation interfaces for real-world applications. For supplementary materials, additional experiments, and qualitative results, visit the project webpage: https://bit.ly/41EfcJa

Figures (4)

• Figure 1: (a) A single SPARK arm with 1:2 scale links, corresponding to the UR5e robotic arm. (b) The force glove worn by an operator interacting with the SPARK end-effector, featuring six motors that relay haptic force directions from the UR5e force/torque sensor.
• Figure 2: Sequence of bimanual operations on all the 5 tasks described in the Table. \ref{['tab:task_table']} are shown here.
• Figure 3: Average task completion times across five tasks. Each task is evaluated 10 times, where the completion time and number of e-stops is collected. E-stops are represented as red triangles. In-Person: The operator can see the manipulators, Remote-Basic: The operator is in a remote location with a 2D video stream, Remote-FG: The operator has a screen and a force glove, Remote-FC: The operator has a screen and the manipulator has force compliance, Remote-FGC: The operator has a screen the force glove and the manipulator has force compliance. Note the use of Remote- instead of SPARK-Remote- for brevity.
• Figure 4: An in-person evaluation compares SPARK with commercial teleoperation solutions, including VR controllers and the SpaceMouse. Results show that SPARK in-person outperforms both alternatives (aligning with findings from GELLO gello). Transitioning to Remote-Basic increases task completion time, while integrating the force glove (FG) and force controller (FC) in Remote-FGC reduces completion time and significantly decreases e-stops, bringing performance closer to in-person SPARK.