Left/Right Brain, human motor control and the implications for robotics

TL;DR

This work investigates whether a biologically inspired bilateral neural architecture, with hemispheric specialization, can improve robotic motor control. By training two hemispheres with different loss functions that mimic dominant (coordinated, efficient movement) and non-dominant (stable, positionally accurate) behaviors, the study demonstrates that specialization can be induced and that bilateral systems, especially with Corpus Callosum coupling, can outperform non-specialized and unilateral baselines on both a random-reach and a hold-position task. The findings show that hemispheres can jointly or independently contribute to task performance, supporting a complementary dominance view and suggesting practical gains for industrial motor control. The approach provides a foundation for designing bilateral AI controllers that leverage hemisphere-specific strengths, with implications for more adaptable and robust robotic manipulation. In essence, specialized bilateral architectures offer a pathway to combine precision and stability in robotics, inspired by human motor control.

Abstract

Neural Network movement controllers promise a variety of advantages over conventional control methods, however, they are not widely adopted due to their inability to produce reliably precise movements. This research explores a bilateral neural network architecture as a control system for motor tasks. We aimed to achieve hemispheric specialisation similar to what is observed in humans across different tasks; the dominant system (usually the right hand, left hemisphere) excels at tasks involving coordination and efficiency of movement, and the non-dominant system performs better at tasks requiring positional stability. Specialisation was achieved by training the hemispheres with different loss functions tailored to the expected behaviour of the respective hemispheres. We compared bilateral models with and without specialised hemispheres, with and without inter-hemispheric connectivity (representing the biological Corpus Callosum), and unilateral models with and without specialisation. The models were trained and tested on two tasks common in the human motor control literature: the random reach task, suited to the dominant system, a model with better coordination, and the hold position task, suited to the non-dominant system, a model with more stable movement. Each system outperformed the non-preferred system in its preferred task. For both tasks, a bilateral model outperformed the non-preferred hand and was as good or better than the preferred hand. The results suggest that the hemispheres could collaborate on tasks or work independently to their strengths. This study provides ideas for how a biologically inspired bilateral architecture could be exploited for industrial motor control.

Figures (7)

• Figure 1: a) MotorNet RigidTendon26 Plant codol_motornet_2023. EE and EF refer to the elbow extensors and flexors, SE and SF to the shoulder extensors and flexors, BE and BF to the bi-articular extensors and flexors. b) Bilateral network with a Corpus Callosum.
• Figure 2: Lesion Positions across a Bilateral Network
• Figure 3: All model performance on both tasks. Each system is trained with a loss function suited to one of the tasks, resulting in a 'preferred' system per task.
• Figure 4: Lesion performance of Bi-S and Bi-NS
• Figure 5: Lesion performance of CC-S and CC-NS.