Real-time reinforcement for human-machine interface control

Pierre Vassiliadis, Daniel Leal Pinheiro, Lisa Fleury, Alexandre Zenon, Martín Esparza-Iaizzo, Abigaïl Ingster, Silvestro Micera, Solaiman Shokur, Friedhelm C. Hummel
Neuron. 2026-06-01; :
DOI: 10.1016/j.neuron.2026.05.009

1. Neuron. 2026 Jun 15:S0896-6273(26)00380-6. doi: 10.1016/j.neuron.2026.05.009.
Online ahead of print.

Real-time reinforcement for human-machine interface control.

Vassiliadis P(1), Pinheiro DL(2), Fleury L(3), Zenon A(4), Esparza-Iaizzo M(3),
Ingster A(3), Micera S(5), Shokur S(6), Hummel FC(7).

Author information:
(1)Defitech Chair of Clinical Neuroengineering, Neuro-X Institute (INX), École
Polytechnique Fédérale de Lausanne (EPFL), 1202 Geneva, Switzerland; Defitech
Chair of Clinical Neuroengineering, INX, EPFL Valais, Clinique Romande de
Réadaptation, 1951 Sion, Switzerland. Electronic address:
.
(2)Bertarelli Foundation Chair in Cognitive Neuroprosthetics, Neuro-X Institute,
École Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland; The
BioRobotics Institute, Health Interdisciplinary Center and Department of
Excellence in Robotics and AI, Scuola Superiore Sant’Anna, 56127 Pisa, Italy.
(3)Defitech Chair of Clinical Neuroengineering, Neuro-X Institute (INX), École
Polytechnique Fédérale de Lausanne (EPFL), 1202 Geneva, Switzerland; Defitech
Chair of Clinical Neuroengineering, INX, EPFL Valais, Clinique Romande de
Réadaptation, 1951 Sion, Switzerland.
(4)INCIA, CNRS, Bordeaux, France.
(5)Bertarelli Foundation Chair in Cognitive Neuroprosthetics, Neuro-X Institute,
École Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland; The
BioRobotics Institute, Health Interdisciplinary Center and Department of
Excellence in Robotics and AI, Scuola Superiore Sant’Anna, 56127 Pisa, Italy;
MINE Laboratory, Università Vita-Salute San Raffaele, Milano, Italy.
(6)Bertarelli Foundation Chair in Cognitive Neuroprosthetics, Neuro-X Institute,
École Polytechnique Fédérale de Lausanne, 1015 Lausanne, Switzerland; The
BioRobotics Institute, Health Interdisciplinary Center and Department of
Excellence in Robotics and AI, Scuola Superiore Sant’Anna, 56127 Pisa, Italy;
MINE Laboratory, Università Vita-Salute San Raffaele, Milano, Italy; Department
of Clinical Neurosciences, University Hospital of Lausanne, University of
Lausanne, Lausanne, Switzerland.
(7)Defitech Chair of Clinical Neuroengineering, Neuro-X Institute (INX), École
Polytechnique Fédérale de Lausanne (EPFL), 1202 Geneva, Switzerland; Defitech
Chair of Clinical Neuroengineering, INX, EPFL Valais, Clinique Romande de
Réadaptation, 1951 Sion, Switzerland; Clinical Neuroscience, University of
Geneva Medical School, 1202 Geneva, Switzerland. Electronic address:
.

A major challenge involved in human-machine interfaces is developing feedback
strategies that improve control and benefit patients with motor disabilities.
Here, we propose, validate, and mechanistically characterize a personalized,
closed-loop strategy that delivers reinforcement feedback in real time during
human-machine interface control. Across five experiments involving 106
participants and two control interfaces, fewer than 20 reinforcement trials
produced immediate improvements in force control and lasting retention gains.
These effects were strongest when visual and/or somatosensory feedback was
limited, a finding that suggests translational relevance for tasks,
technologies, and pathologies with limited sensory feedback. In chronic stroke
patients, real-time reinforcement likewise improved online force control under
limited visual feedback, although short training did not yield retention gains.
Information-theoretic analyses further revealed that reinforcement compensates
for reduced feedback control when sensory feedback is sparse and promotes motor
exploitation of successful actions. Overall, these findings identify real-time
reinforcement as a promising strategy for enhancing human-machine interface
control.