HaptiNet: Networked Haptic Robots Enable Physical Co-presence in Geographically-Unconstrained Rehabilitation

• Neurological impairment is linked to social isolation, while conventional rehabilitation remains predominantly individualistic and limited in social interaction.Source: Main Text: INTRODUCTION · p. 3
• Remote cooperative rehabilitation is constrained by the need to transmit forces and coordinate movements despite geographic separation and network latency.Source: Main Text: INTRODUCTION · p. 3
• HaptiNet combines low-inertia, long-stroke, high-force-feedback terminals with a shared virtual environment and imitation-learning motion prediction.Source: Main Text: INTRODUCTION · p. 4
• Across 284 healthy participants and 111 patients, HaptiNet preserved force transmission and improved task performance, engagement, and interpersonal motor synchrony.Compared with solo training, reported improvements were 24% in task performance, 49% subjective engagement, 27% objective engagement, and 16% motor synchrony.Source: Main Text: INTRODUCTION · p. 4
• Across single-user, two-user, and three-user conditions, task-required force remained consistent under 4 N, 10 N, and 20 N resistance levels.Trial-averaged absolute force errors remained within 4%, 2%, and 3%, respectively.Source: HaptiNet preserves haptic-rendering… · p. 5

• The experiment separated benefits of haptic interaction from those of multi-user participation and shared visual feedback alone.Source: HaptiNet-mediated haptic coordination… · p. 5, Figure 1 · p. 6
• Haptic co-op produced higher game scores and lower tracking errors than both Solo and Visual co-op.Game-score comparisons were significant versus Solo (P = 0.001) and Visual co-op (P = 0.005); tracking-error comparisons were P < 0.001 for both.Source: HaptiNet-mediated haptic coordination… · p. 5
• Physical effort was greater in both multi-user conditions than in Solo, indicating that participation increased motor involvement.Source: HaptiNet-mediated haptic coordination… · p. 6
• 4.59 ± 0.52 was the engagement rating for Haptic co-op, versus 3.63 ± 0.57 for Solo and 3.71 ± 0.65 for Visual co-op.Source: HaptiNet-mediated haptic coordination… · p. 6
• Haptic co-op increased haptic realism and cooperation realism relative to Visual co-op while maintaining haptic realism comparable to Solo.Haptic realism was 4.46 ± 0.49 versus 3.73 ± 0.81, and cooperation realism was 4.70 ± 0.44 versus 3.86 ± 0.91.Source: HaptiNet-mediated haptic coordination… · p. 6

• EEG measured physiological engagement while sEMG measured interpersonal action synchrony during Solo and Haptic cooperative task performance.Source: Haptic interaction enhances… · p. 7, Figure 2 · p. 9
• Normalized engagement index increased during Co-op at F3, F4, and their bilateral average across the 3–13 s task interval.Values changed from 0.93 to 1.06 at F3, 0.94 to 1.06 at F4, and 0.94 to 1.06 for the bilateral average.Source: Haptic interaction enhances… · p. 7
• Co-op increased interparticipant sEMG coherence across multiple upper-limb muscles compared with Solo.In the 1–3 Hz band, coherence was consistently higher in six of eight analyzed muscles.Source: Haptic interaction enhances… · p. 8
• The combined EEG and sEMG results showed enhanced physiological engagement and interpersonal motor synchrony during haptic cooperation.Source: Haptic interaction enhances… · p. 8

• Task score decreased and resultant force increased as latency rose from 0 ms to 50 ms.The tested conditions were 0 ms, 10 ms, 30 ms, and 50 ms, corresponding to local, LAN, MAN, and WAN communication.Source: Robust haptic interaction under… · p. 10
• HaptiNet compared no prediction, constant-velocity prediction, and imitation-learning prediction to compensate for delayed remote motion.Source: Robust haptic interaction under… · p. 10, Figure 3 · p. 11
• At 50 ms, ILMP more closely followed the true remote trajectory and suppressed reversal-related overshoot compared with NMP and CVMP.Source: Robust haptic interaction under… · p. 10
• 0.193, 0.051, and 0.009 were the latency-dependent force-ratio slopes for NMP, CVMP, and ILMP, respectively.ILMP kept resultant force closest to the no-latency baseline as latency increased.Source: Robust haptic interaction under… · p. 10
• At 30 ms, ILMP outperformed NMP on task score, and at 50 ms it outperformed both NMP and CVMP.At 50 ms, ILMP exceeded CVMP with P = 0.011.Source: Robust haptic interaction under… · pp. 10–11

• Three intercity links exhibited approximately 34.5, 40.1, and 43.8 ms delays for A–B, A–C, and B–C connections, respectively.Source: Long-distance deployment of HaptiNet… · p. 12
• Co-op 2 using ILMP achieved the highest task scores during both training and post-training testing.Co-op 2 significantly outperformed Solo and Co-op 1 in both phases (P < 0.001 for each comparison).Source: Long-distance deployment of HaptiNet… · p. 12
• ILMP preserved engagement and haptic realism better than CVMP during long-distance cooperation.Co-op 2 produced higher engagement than Solo and Co-op 1, while Co-op 1 had lower haptic realism than Solo and Co-op 2.Source: Long-distance deployment of HaptiNet… · pp. 12–13
• ILMP preserved resultant-force rendering at a level not significantly different from Solo, whereas CVMP differed significantly from both conditions.Source: Long-distance deployment of HaptiNet… · p. 13
• The remote protocol included baseline solo testing, randomized Solo and cooperative conditions, post-training testing, questionnaires, and force analysis.Source: Figure 4 · p. 13

• 111 patients with neurological impairments participated in solo or HaptiNet-mediated cooperative rehabilitation, including local and remote sessions.Most participants were post-stroke patients; other conditions included brain tumors, moyamoya disease, and related disorders.Source: Clinical validation in patients with… · p. 14
• Co-op score improvement was 12.81 versus 3.32 for Solo, while tracking MAE reduction was 0.88 cm versus 0.33 cm.Baseline scores and tracking MAE did not differ significantly between groups.Source: Clinical validation in patients with… · pp. 14–15
• Co-op produced larger improvements than Solo in both task score and tracking accuracy as training progressed.Source: Clinical validation in patients with… · p. 15
• The system architecture combined distributed low-inertia, long-stroke terminals with imitation-learning-based motion prediction for delay-compensated cooperation.Source: Clinical validation in patients with… · p. 16
• Validation progressed from fundamental haptic rendering and transmission tests to network-delay experiments and clinical rehabilitation scenarios.Source: Clinical validation in patients with… · pp. 16–17
• In the clinical stage, Co-op improved from baseline by 23% versus 6% for Solo and produced 106% greater patient-applied effort.The Co-op improvement was approximately 3.87-fold greater than Solo.Source: Clinical validation in patients with… · p. 17
• The results mainly reflect immediate training improvements and should not be equated directly with long-term motor functional recovery.The authors call for larger cohorts, longer interventions, and more rigorous randomized controlled designs.Source: Clinical validation in patients with… · p. 17
• The study supports HaptiNet as a platform for cooperative rehabilitation beyond co-located settings.Source: Clinical validation in patients with… · p. 17

• Each terminal conveys a user’s interaction forces and motion to other connected users through four integrated architectural components.Source: MATERIALS AND METHODS System Development · p. 18

Robotic device

• The robotic device uses a jointly optimized transmission mechanism and motion range to combine low inertia, high force output, and lightweight construction.Timing-belt transmission was selected to reduce reflected inertia and support a lightweight structure.Source: MATERIALS AND METHODS System Development · p. 18

Control framework for local haptic rendering

• The local haptic control framework supports no-load, single-user loaded, and multi-user cooperative task modes.Source: MATERIALS AND METHODS System Development · p. 18

Shared virtual environment synchronization module

• The shared virtual environment synchronizes local replicas of the task object and remote partner proxies across terminals.Source: MATERIALS AND METHODS System Development · p. 18

ILMP-based delay compensator

• The ILMP-based delay compensator predicts current remote-partner states from delayed motion information to reduce latency-induced motion asynchrony.Source: MATERIALS AND METHODS System Development · p. 18

• Each terminal maintains local representations of all users’ virtual proxies and a dynamic replica of the shared task object.Source: Shared virtual environment… · p. 19
• Virtual coupling converts relative proxy–object motion into interaction forces applied to each local object replica.The coupling uses spring and damping coefficients.Source: Shared virtual environment… · p. 19
• The task-level interaction force is computed by summing the virtual-coupling forces generated by all users.Source: Shared virtual environment… · p. 19
• Synchronization coupling reduces state deviations among object replicas caused by network delay, packet loss, prediction error, and numerical integration error.Its correction force is normalized by 1/(n−1) across remote replicas.Source: Shared virtual environment… · pp. 19–20
• The synchronization correction maintains consistency without being interpreted as a task-level user interaction force.Object motion is primarily determined by the combined virtual-coupling forces, while synchronization preserves shared-state consistency.Source: Shared virtual environment… · p. 20

• ILMP learns task-related motion patterns from recorded user–object interactions and estimates future velocity trajectories from recent motion histories.Source: Model training for ILMP · p. 20
• Single-user loaded movement tasks provide training samples linking user motion, virtual object motion, and target trajectory information.Motion data are resampled to the 100 Hz network-update rate; the local control loop runs at 1000 Hz.Source: Model training for ILMP · p. 20
• Two independent MLP predictors estimate remote-partner motion and shared-object motion.Source: Model training for ILMP · p. 20
• The user-motion predictor uses position and velocity histories, box-position history, and the target-position sequence as inputs.Source: Model training for ILMP · p. 20
• Each predictor minimizes mean squared error between predicted and measured future velocity sequences.The predictors separately represent user-velocity and box-velocity prediction.Source: Model training for ILMP · p. 21

• At each control cycle, ILMP uses the latest received remote position and velocity to predict a short-horizon future velocity sequence.The study outputs ten predicted velocity samples.Source: Online motion prediction · p. 21
• The predicted velocity sequence is interpolated from the 100 Hz network-update rate to the 1000 Hz local robot-control rate.Source: Online motion prediction · p. 21
• The latency-compensated remote position is reconstructed by integrating interpolated predicted velocity from the latest received position.Source: Online motion prediction · p. 21
• The compensated state corresponding to measured network latency is used first, followed by subsequent predictions until new network information arrives.The reconstructed position updates the local replica of the remote user or shared object for synchronization and haptic rendering.Source: Online motion prediction · p. 21

• The study was approved by the Ethics Committee of Southern University of Science and Technology under approval number 20230095.Source: Participants and experimental protocol · p. 21
• Participants included 284 healthy individuals and 111 patients with neurological impairments.Source: Participants and experimental protocol · pp. 21–22
• All participants provided written informed consent before the experiments.Source: Participants and experimental protocol · p. 22

• Statistical analyses used MATLAB R2025b.Source: Statistical analysis · p. 22
• Detailed statistical methods and results are available in the Supplementary Materials.Source: Statistical analysis · p. 22
• A more detailed statistical report and raw data are provided as accompanying CSV files.Source: Statistical analysis · p. 22