A Reliable EMG-Based Robotic Interface for Assistive Mobility with Sensor-Driven Safety Control

This work set out to demonstrate that a differential-drive mobility platform could be reliably governed by forearm EMG signals using threshold logic alone, without resorting to gesture classification, machine learning, or per-user calibration — and the experimental evidence gathered here supports that premise. Dual-channel acquisition through Muscle BioAmp Candy sensors, conditioned via a fourth-order Butterworth band-pass filter and reduced to a stable envelope through circular-buffer averaging, produced a clear and consistent separation between resting (~14) and active (~65) signal states across all three test subjects — a margin wide enough to accommodate the modest inter-subject and inter-session variability observed during testing without any threshold retuning. Equally significant is the performance of the embedded safety architecture. By treating ultrasonic obstacle sensing and infrared edge detection as hard constraints on motor activation rather than advisory inputs, the system achieved zero false negatives across all hazard trials conducted, validating the design decision to prioritize a conservative, fail-to-stop default over continuous operation. This outcome directly addresses the safety-deployment gap identified earlier in the literature, where control sophistication alone has been shown insufficient to guarantee real-world reliability in assistive robotic platforms. Taken together, the results indicate that threshold-based EMG control, when paired with deliberate signal conditioning and a hardware-level safety interlock, offers a practically viable and economically accessible pathway toward assistive mobility — one whose simplicity is not a compromise but rather a design choice suited to the embedded, low-cost context this work specifically targets. Future development directions, including adaptive threshold calibration, proportional speed control derived from envelope amplitude, and expanded lateral hazard sensing, are positioned to extend this platform's applicability toward a broader and more clinically representative user population.

  [1]     A. Nanavati, V. Ranganeni, and M. Cakmak, "Physically assistive robots: A systematic review of mobile and manipulator robots that physically assist people with disabilities," Annual Review of Control, Robotics, and Autonomous Systems, vol. 7, 2024. [2]     "EEG and EMG-based human-machine interface for navigation of mobility-related assistive wheelchair (MRA-W)," Heliyon, vol. 10, no. 5, Mar. 2024, Art. no. e27777, DOI: 10.1016/j.heliyon.2024.e27777. [3]     S. Bisi, L. De Luca, B. Shrestha, and Z. Yang, "Development of an EMG-controlled mobile robot," Robotics, vol. 7, no. 3, pp. 36, Jul. 2023. [4]     G. D'Onofrio, L. Fiorini, H. Hoshino et al., "Assistive robots for socialization in elderly people: Results pertaining to the needs of the users," PubMed, 2023. [5]     A. C. Manero, S. L. McLinden, J. Sparkman, and B. Oskarsson, "Evaluating surface EMG control of motorized wheelchairs for amyotrophic lateral sclerosis patients," Journal of NeuroEngineering and Rehabilitation, vol. 19, no. 1, 2022, DOI: 10.1186/s12984-022-01066-8. [6]     F. Grzeskowiak et al., "SWALKIT: A generic augmented walker kit to provide haptic feedback navigation assistance to people with both visual and motor impairments," in 2022 International Conference on Rehabilitation Robotics (ICORR), IEEE, 2022, DOI: 10.1109/ICORR55369.2022.9896406. [7]     R. A. R. C. Gopura, S. Bandara, and M. P. Gunasekara, "Recent trends in EMG-based control methods for assistive robots," in Electrodiagnosis in New Frontiers of Clinical Research, InTech, DOI: 10.5772/56174. [8]     World Health Organization, Disability – Fact Sheet, WHO Report, Geneva, 2023.