Browse nine paper contributions and two demonstrations, with manuscripts and demo videos. Explore the contributions →
Human-robot interaction (HRI) is moving beyond conventional physical interfaces and isolated feedback channels toward sensorimotor augmentation that more closely matches human perception, proprioception, and decision-making. Despite major advances in robot learning, multimodal sensing, and shared autonomy, important challenges remain in reducing operator workload, preserving situational awareness, and enabling intuitive and safe HRI under uncertainty.
The workshop expands the concept introduced in our recent ICRA/IROS 2025 events by focusing on three concrete advances. First, it examines digital-twin-enhanced training and evaluation, including simulation-based scenario design, sim-to-real calibration, and demonstration data collection for reproducible benchmarking. Second, it addresses adaptive shared autonomy with measurable safeguards, including authority allocation, risk-sensitive control, safety filters, and online adaptation to operator state and task complexity. Third, it broadens the application scope beyond teleoperation in the narrow sense to include interactive assistance, rehabilitation, surgical robotics, marine operations, and nuclear decommissioning, where human capability, environmental uncertainty, and safety constraints must be treated jointly.
Unlike previous workshops focused on unilateral teleoperation, this event uniquely explores sensorimotor integration across diverse applications - surgery, marine exploration, and nuclear decommissioning - where human adaptability and real-time feedback are critical. Building upon the success of the 4th edition at IROS 2025 in Hangzhou, this 5th edition introduces new themes and an expanded half-day format. Topics include:
The half-day workshop features 5 confirmed speakers from academia and industry, an innovative Mentor-Mentee Network (MMN) fostering cross-disciplinary collaboration, interactive poster and demonstration activities, and an industry showcase. By integrating robotics, neuroscience, and human-computer interaction, this workshop seeks to reconceptualize teleoperation through human-centered sensorimotor enhancements, prioritizing safety and intuitive interaction to enable next-generation collaborative frameworks.
Human-Centric Multilateral Teleoperation Augmentation: Perception, Telecommunication, and Coordination
Multisensory Transparency-Augmented Teleoperation in Extreme Environments
Real World Physical and Social Human-Robot Interaction
Human-Centric Multilateral Teleoperation
4th Sensorimotor-Augmented Teleoperation: HIO Paradigm and Real-World Concerns
Photos from IROS 2026 will be posted here after the event. Below are highlights from our previous workshops.
The workshop is a half-day afternoon event. All times are in Eastern Time (Pittsburgh local time).
| Time | Talk | Comments |
|---|---|---|
| 13:30 - 13:40 | Welcome and Workshop Introduction — Dr. Ziwei Wang | Workshop overview, motivation, and introduction to HIO paradigm |
| Session 1: Adaptive and Multimodal Human–Robot Interaction | ||
| 13:40 - 14:05 | Prof. Dongjun Lee — How to Intuitively and Efficiently Teleoperate a Limited-Dexterity Robot Hand for Contact-Rich Manipulation Data Collection | Invited talk |
| 14:05 - 14:30 | Prof. Hao Su — Deep Reinforcement Learning in Simulation for Versatile Wearable and Surgical Robots | Invited talk |
| 14:30 - 14:45 | Moderated Q&A and Cross-Talk | Discussion across control, assistive robotics, and wearable interaction |
| 14:45 - 15:15 | Prof. Chenguang Yang — Skill-Based Learning and Compliant Control for Enhanced Human-Robot Teleoperation and Interaction | Invited talk |
| 15:15 - 16:00 | Coffee Break, Posters and Demonstrations | Official afternoon break; informal discussion around posters and demo systems |
| Session 2: Sensorimotor Augmentation in Real-World Teleoperation | ||
| 16:00 - 16:25 | Dr. Enrica Zereik — Embodied Intelligence and Social Cues for Cooperation in Multi-Robot Systems | Invited talk |
| 16:25 - 16:50 | Prof. Ya-Jun Pan — Effective Adaptive Robot-robot and Human-robot Interaction | Invited talk |
| 16:50 - 17:25 | Poster Session | Selected posters and demos |
| 17:25 - 17:30 | Closing Remarks and Next Steps | Summary of technical challenges, benchmarking needs, and deployment priorities |
Workshop contributions: Browse papers · Watch demos
We thank our sponsors, technical committee partner, and journal partner for supporting the workshop.
JATEN Robotics develops and manufactures AGV/AMR mobile robots and integrated intelligent-logistics systems for industrial material handling. Founded in 2002, the company provides mobile-robot platforms and factory-wide logistics solutions. Visit JATEN Robotics
ChangingTek develops dexterous robotic hands and embodied-intelligence robot technologies, integrating adaptive mechanisms, multimodal sensing, and AI-based control for industrial, research, and emerging embodied-AI applications. Visit ChangingTek
Robot Learning is an ELSPublishing journal (ISSN 2960-1436 print; 2960-1444 online). Selected workshop contributions may be considered for a journal transfer opportunity, subject to the journal’s scope, editorial assessment, and peer-review process. Journal information
The IEEE Industrial Electronics Society Technical Committee on Control, Robotics, and Mechatronics brings together researchers and practitioners working in robot control, telemanipulation, telerobotics, telepresence, intelligent spaces, and emerging mechatronics technologies. Visit TC-CRM
Explore the workshop contributions, including nine papers and two demonstrations.
Northeastern University, China
An execution-aware framework that separates the arrival of a navigation reference from its acceptance by a moving humanoid. It adapts reference geometry and compares keeping, updating, or transitioning between trajectories, with evaluation in simulated Unitree G1 stair navigation.
Shanxi University, China
A physics-grounded RGB-D grasp detection framework that extracts illumination-invariant structural cues for low-light manipulation. The study introduces the RLL-Grasp dataset and reports evaluations on the dataset and in real-world robotic grasping trials.
Lancaster University; Tangi0 Ltd (TG0), United Kingdom
A wearable framework that combines plantar pressure and inertial measurements using spatial and temporal attention to estimate three-dimensional ground reaction forces. Evaluation against a force plate in five healthy adults reports improved performance over a dual-stream attention baseline, with broader validation remaining future work.
Nantong University, China
A comparison of synchronous, geometry-guided, and content-controlled latent generation under matched sampling budgets. Geometry-guided progression improves latent reconstruction metrics, while the tested content signals do not establish an overall advantage over the geometric prior.
Lubbock High School; NYU Tandon School of Engineering, USA
A study of training augmentation and prior-frame context for catheter and guidewire segmentation on the CathAction human subset. On the matched test frames, neither temporal model improves mean Dice over the augmented single-frame U-Net; robotic control and clinical performance were not evaluated.
Beijing University of Technology; The Hong Kong Polytechnic University; SynnoverAI; Technical University of Munich; University of Hamburg; Eastern Institute of Technology
A leader–follower teleoperation method that accounts for coupled transmissions, gravity loads, and actuator saturation. Coordinated motor-space saturation and feasible-reference updates preserve operator motion intent and improve trajectory continuity in real-robot experiments.
Lancaster University, United Kingdom
A proposed gaze-controlled dual-arm shared autonomy framework for assistive manipulation in constrained elderly-care environments. A visual separation scheme allocates attention between command generation and situational monitoring; the paper outlines evaluation tasks and expected outcomes.
Lancaster University, United Kingdom; Tsinghua University, China
A proposed dual-layer reinforcement learning framework combining patient-specific gait modeling with anticipatory exoskeleton assistance. The approach integrates musculoskeletal simulation, progressive rehabilitation rewards, and instrumented insole sensing, with validation planned from simulation to pilot hardware deployment.
Xi’an Jiaotong-Liverpool University, China
A residual-control execution layer for retaining one object while grasping another with a dual-sided dexterous hand. A fixed upstream plan is combined with learned retention adjustments and deterministic command constraints; simulation results provide preliminary feasibility evidence.
Watch the demonstration videos and read the accompanying submissions.
Case Western Reserve University, USA
A SurRoL-based peg-transfer simulator with bimanual Touch-device teleoperation, stereo display, and Varjo eye tracking. The demo presents text, point-of-interest, and foveation guidance under consistent task conditions; user-performance evaluation is future work.
Harbin Institute of Technology, Shenzhen; South China University of Technology, Guangzhou, China
A four-camera tactile-sensing prototype that folds optical paths over an 86 mm × 86 mm silicone contact surface. Shared image streams support shape classification, contact-trajectory recognition, and multiview deformation reconstruction, with preliminary evidence from one prototype.
All papers must be submitted in PDF format following the IROS double-column format. Templates are available at the IEEE templates page. Accepted papers will be presented as posters. Selected papers featured in the Poster Spotlight initiative. The workshop is non-archival.
Submission: Please submit via email to luzhenyu@scut.edu.cn with the subject line IROS 2026 WS Submission.
Workshop Best Poster Award (US$500) and Workshop Best Demo Award (US$500) will be presented at the workshop. Selected workshop contributions may be invited to submit substantially extended manuscripts to the following journal special issues or transfer option. All journal submissions are subject to the journals' scope, editorial assessment, and standard peer-review process. Extended submission deadline for all journal opportunities: November 30, 2026.
The MMN pairs early-career researchers with established experts for pre-workshop collaborative projects. Selected pairs showcase their work during the afternoon MMN Demonstration Session. Contact luzhenyu@scut.edu.cn to join as a mentor or mentee.
IROS 2026 will be held at the David L. Lawrence Convention Center in Pittsburgh, Pennsylvania, USA. Pittsburgh, known as Roboburgh, has transformed from the historic Steel City into a global hub for robotics and AI, home to Carnegie Mellon University's Robotics Institute and the National Robotics Engineering Center.
David L. Lawrence Convention Center
1000 Fort Duquesne Blvd, Pittsburgh, PA 15222, USA
Official IROS listing: Half Day · 1:30 PM–5:30 PM · Room 415. Use the official IROS link above for the conference room map.
Pittsburgh International Airport (PIT) is approximately 30 minutes from the Convention Center. Taxi, rideshare (Uber/Lyft), and the 28X Airport Flyer bus provide direct connections to downtown.
Hotels within walking distance include Westin Pittsburgh, Omni William Penn, and Drury Plaza. Check the IROS 2026 website for conference rates.
Pittsburgh has bus and light rail (T) via PRT. The Convention Center is near Wood Street and Steel Plaza T stations. Downtown is walkable.
Home to CMU Robotics Institute, NREC, Aurora Innovation, and numerous startups. A city where cutting-edge robotics research meets real-world innovation.
