Toward Humanoid Robots in Construction: A Teleoperation Feasibility Study

Parastoo Ali Pour; David R. Martin; Chang Min Hur; Bo Zhang; Tommy Zhou; Brandon Thomas Lichter; Shane Stanfield; Pramod Khargonekar

cs.ROSep 30, 2026

Toward Humanoid Robots in Construction: A Teleoperation Feasibility Study

Authors: Parastoo Ali Pour, David R. Martin, Chang Min Hur, Bo Zhang, Tommy Zhou, Brandon Thomas Lichter, Shane Stanfield, Pramod Khargonekar, +1 more

Organizations: Department of Mechanical and Aerospace Engineering, University of California at Irvine, Irvine, CA 92697 USA · Department of Electrical Engineering and Computer Science, University of California at Irvine, Irvine, CA 92697 USA · Professional Master of Embedded and Cyber-physical Systems Program, University of California at Irvine, Irvine, CA 92697 USA

Abstract

We present a teleoperation system that enables a single operator to perform construction tasks on a Unitree G1 humanoid, combining extended reality (XR) based upper body control with pedal-based locomotion to enable simultaneous manipulation and locomotion. Motivated by persistent labor shortages, hazardous working conditions, and challenges in humanoid autonomy, we investigate teleoperation as a practical near-term approach for reducing physical strain on workers while generating high quality demonstration data. We evaluate the system on two representative construction tasks drawn from O*NET occupational database, and report task success and completion time relative to a manual baseline. The system achieved 100% success on tool transport and 80% success on surface painting, with teleoperation requiring substantially more time compared to manual execution.

Figures & tables

Fig. 1: Humanoid teleoperation for construction task. (Left) An operator controls the robot using a Meta Quest 3 headset and a GYLDR foot controller. (Right) A teleoperated Unitree G1 performs a tool transport task on a job site. It grasps a hand tool (a trowel) from the work surface and carries it toward the destination container (red box).

Fig. 2: Diagram of the teleoperation system architecture. The operator controls the Unitree G1 humanoid using a Meta Quest 3 headset for upper-body and neck motion and GLYDR foot controller for locomotion. The control inputs are processed through a controller PC, while egocentric video from the robot-mounted ZED Mini camera is streamed to the operator. All camera, tactile, motor commands, and velocity data are recorded in LeRobot format to support future policy learning.

Task	O*NET Category	Description
1	Core	Load, unload, or identify building materials, machinery, or tools, distributing them to the appropriate locations, according to project plans or specifications.
2	Supplemental	Mop, brush, or spread paints, cleaning solutions, or other compounds over surfaces to clean them or to provide protection.

TABLE I: Selected Tasks from O*NET Occupation Specific Tasks for Construction Laborers [ 13 ]

Task	Operator	Teleoperation Success Rate (%)	Avg. Teleoperation Time (s)	Avg. Baseline Time (s)	Trials
Tool Transport	VR-novice	100	83	5	10
Tool Transport	VR-expert	100	62	3	10
Tool Transport	Avg	100	72	4	20
Painting	VR-novice	80	169	40	10
Painting	VR-expert	80	129	53	10
Painting	Avg	80	149	47	20

TABLE II: Teleoperation task performance across two representative tasks.

Explore similar work

Perception-and-action system for humanoid robot task execution in construction
Aug 3, 2026Yanxi Liu, Yizhi LiuPerception Pipeline Construction
Department of Civil and Environmental Engineering, Syracuse University, 158 Link Hall, Syracuse, NY 13244, USA · Department of Civil and Environmental Engineering, Syracuse University, 151J Link Hall, Syracuse, NY 13244, USA
Immersive Social Interaction with VR and LLM-Assisted Humanoids
Jul 8, 2026Niraj Pudasaini, Geeta Chandra Raju Bethala, Pranav Doma +2Teleoperation Full-Body Humanoid Control
New York University Abu Dhabi (NYUAD), UAE
Teleopit: A Full-Embodiment Humanoid Teleoperation System
Aug 3, 2026Bingqian Wu, Zicheng Xu, Xianghui Fan +2Teleoperation Cross-Embodiment Retargeting
Westlake University · Shanghai Innovation Institute

Aug 3, 2026, cs.ROJ/K move · Enter open · S save

Perception-and-action system for humanoid robot task execution in construction

Yanxi Liu, Yizhi Liu

Department of Civil and Environmental Engineering, Syracuse University, 158 Link Hall, Syracuse, NY 13244, USA · Department of Civil and Environmental Engineering, Syracuse University, 151J Link Hall, Syracuse, NY 13244, USA

Perception Pipeline Construction Deep Network

Open paper AskPDF arXiv

Humanoid robots, with their human-like shape and multi-tasking capabilities, are well-aligned with human-dominated workplaces, like those in civil and construction engineering, where they could collaborate with human workers or autonomously perform physically demanding and hazardous tasks. Despite this promise, limited research has explored how to endow these robots with the practical capabilities needed to perform construction tasks. To this end, this study proposes a novel perception-and-action system that enables humanoid robots to learn and perform construction tasks from worker demonstrations. This system contains two deep networks: Humanoid-PoseNet, which extracts human postures and translates them into mechanically feasible poses for a humanoid robot; and Humanoid-ActionNet, which learns robot-executable actions based on these translated poses. Experimental results demonstrate that the humanoid robot reliably executed eight construction-related actions, achieving an average motion-tracking error of 82.45 mm MPJPE (Mean Per Joint Position Error). This work provides an early step toward deploying humanoid collaborators in construction.

Abstract

Explore similar work

Perception-and-action system for humanoid robot task execution in construction

Immersive Social Interaction with VR and LLM-Assisted Humanoids

Teleopit: A Full-Embodiment Humanoid Teleoperation System