Upper-Back Vibrotactile Feedback Holds Up During Reaching, While the Hand Pays a Motor Cost
Wearable vibrotactile stimulation is attractive because it can deliver feedback without tying up vision or speech. In this preprint, that promise survived a realistic dual-task test: participants could still reach while detecting tactile signals, and stimulation on the upper back interfered less with movement than stimulation on the hand.
The work sits at the intersection of motor control, human factors, and assistive technology. It asks a deceptively simple question with direct design consequences: where should a wearable device deliver vibration if the goal is to support posture, balance, rehabilitation, or skill training without slowing a voluntary movement?
Detection Drops During Movement, but the Body Surface Matters
Participants completed three tasks: a detection task, a reaching task, and a combined reach-and-detect task. Vibrotactile signals were delivered either on the hand or on the upper back, allowing the investigators to compare sensory sensitivity, motor performance, and cognitive workload across body sites and task conditions.
Sensitivity to vibration fell during movement, which is consistent with well-known tactile suppression during action. The decline was not uniform across the body. Suppression was stronger when stimulation was applied to the hand than when it was applied to the upper back, suggesting that tactile signals competing with the efferent and afferent machinery of goal-directed reaching are more vulnerable when delivered to the limb actively engaged in the task.
The result is not trivial for device design. A signal on the hand sits in a crowded sensorimotor channel, where attention is already allocated to the moving effector and where the nervous system may dampen predictable somatosensory input during action. The upper back appears to offer a more forgiving surface for feedback, likely because it is less directly entangled with the mechanics of the reach itself.
Reaching Slowed on the Hand, Not on the Upper Back
Movement time told the clearest practical story. Vibration on the hand slowed reaching, while vibration on the upper back did not impede reaching performance.
This divergence matters because it separates sensory detectability from motor interference. A device can fail in two different ways: the user may miss the signal, or the signal may degrade the primary action the person is trying to complete. Here, upper-back stimulation preserved the reaching task even as detection remained subject to some suppression. Hand stimulation, by contrast, imposed a measurable motor cost. For applications that depend on uninterrupted movement — ergonomic coaching, rehabilitation practice, or closed-loop biofeedback during functional tasks — that asymmetry makes the body location central to the engineering problem.
The physiology is easy to infer even without overreaching beyond the data. The hand is not just a site of sensation; it is the endpoint of the movement. Mechanical vibration there can compete with task-relevant proprioceptive and tactile inputs, nudging the control system into a less efficient mode. On the upper back, the same signal remains accessible to perception but is less likely to hijack the fine motor demands of reaching.
Dual-Tasking Raised Mental Load Regardless of Where the Signal Was Delivered
Cognitive cost did increase when participants had to reach and detect at the same time. Mental workload and task complexity were higher for the combined reach-and-detect condition than for either task alone.
That finding draws a useful boundary around the promise of wearable biofeedback. Upper-back stimulation may spare motor performance, but the user still has to interpret the cue while acting. In other words, preserving movement does not mean the system is effortless. The brain must still divide attention between the reaching goal and the tactile message, and that division carries a subjective burden.
For rehabilitation and motor learning programs, this tradeoff is familiar. A feedback channel can be mechanically unobtrusive yet cognitively demanding, especially when the signal must be noticed quickly or interpreted under time pressure. The study points to a design challenge rather than a failure of the concept: tactile cues on the upper back may need to be highly salient, individualized, or tuned to the user’s perceptual threshold if they are to outperform simpler feedback strategies.
Why the Upper Back Looks Better Suited to Wearable Feedback
The upper back may offer a practical compromise between accessibility and interference. It is large enough to host a wearable device, distant from the hand’s primary role in object-directed action, and capable of carrying a clear tactile cue without the same degree of movement-related disruption seen at the limb.
That is especially relevant for technologies aimed at posture or balance, where the cue does not need to come from the same body segment that is executing the task. In those settings, the feedback channel can be separated from the motor channel. Separating them appears to help.
The authors also point to a possible solution for the residual suppression seen on the upper back: stronger or personalized vibration. If a cue must compete with ongoing movement, salience becomes part of the therapeutic or ergonomic design. A signal that is too weak may disappear into the background of action; one that is more conspicuous may survive dual-task conditions without forcing the user to abandon natural movement.
A Preprint With Clear Design Implications, but Still Early Evidence
This is a preprint, so the findings should be treated as early evidence rather than settled practice guidance. The work does not establish clinical benefit for any disease state, and it does not yet answer how these devices behave across older adults, patients with neurological disease, or people with impaired sensation.
Still, the mechanistic direction is clear. Upper-back vibrotactile feedback appears better tolerated during goal-directed reaching than hand-based stimulation, with less disruption of movement and only minor tactile suppression. The remaining challenge is not whether the concept can work, but how to make the cue salient enough to be useful without adding mental strain.
Future validation will need to test that balance in populations where wearable feedback is most likely to matter: rehabilitation, postural training, and motor learning under real-world conditions. Until then, this study provides a useful engineering rule of thumb — keep the tactile signal off the moving hand when possible, and design the vibration to be strong enough to survive the noise of action.