Objective: Virtual environments provide a safe and accessible way to test innovative technologies for controlling wearable robotic devices for assisting human movement. However, to apply them to systems that support walking, such as powered prosthetic legs, it is not enough to model the hardware itself. Predictive locomotion synthesizers can generate the movements of a virtual user, with whom the simulated device can be trained or evaluated. Methods: We implemented a Deep Reinforcement Learning based motion controller in a physics engine, where autonomy over the humanoid model is shared between the simulated user and the control policy of an active prosthesis. A data-driven, continuous representations of user intent was used to simulate a Human Machine Interface that controlled a transtibial prosthesis. The system was tested in a complex non-steady-state locomotion task, involving turns and stops. Results: Providing the intent to the device control policy did not improve performance of the human-prosthesis system if both policies were learnt simultaneously. However, if the human walking policy was frozen, the intent-driven prosthesis outperformed its counterpart at non-cyclic gait patterns. Conclusion: The continuous intent representation used was shown to mitigate the need for compensatory gait patterns from their virtual users. Co-adaptation was identified as a potential challenge for training prosthesis control policies with human-in-the-loop. Significance: The proposed framework outlines a way to explore the complex design space of robot-assisted gait, promoting the transfer of the next generation of intent driven controllers from the lab to real-life scenarios.

For the control of wearable robotics, it is critical to obtain a prediction of the user’s motion intent with high accuracy. Electromyography (EMG) recordings have often been used as inputs for these devices, however bipolar EMG electrodes are highly sensitive to their location. Positional shifts of electrodes after training gait prediction models can therefore result in severe performance degradation. This study uses high-density EMG electrodes to simulate various bipolar electrode signals from four leg muscles during steady-state walking. The bipolar signals were ranked based on the consistency of the corresponding EMG envelope’s activity and timing across gait cycles. The locations were then compared by evaluating the performance of an offline Temporal Convolutional Network (TCN) that mapped EMG signals to knee angles. The results showed that electrode locations with consistent EMG envelopes resulted in greater prediction accuracy compared to hand-aligned placements (p<0.01). However, performance gains through this process were limited, and did not resolve the position shift issue. Instead of training a model for a single location, we showed that randomly sampling bipolar combinations across the high-density EMG grid during training mitigated this effect. Models trained with this method generalised over all positions, and achieved 70% less prediction error than location specific models over the entire area of the grid. Therefore, the use of high-density EMG grids to build training datasets could enable the development of models robust to spatial variations, and reduce the impact of muscle-specific electrode placement on accuracy.