Research Article
- Volume 2, Issue 1 2026
By Misbah Anwer, Muhammad Fahad, Anusha Hasan, Abdul Karim Hasan, Falak Shah, Rafia Khan
10.21621/ijec.20260201.03
Keywords: sEMG, TinyML, Prosthetics, Edge Computing, Random Forest, ESP32, Incremental Learning, Z-Transform, Signal Processing, 7-DOF

Upper-limb amputations are a significant source of disability, especially in resource-limited environments where expensive commercial-grade prosthetics (often costing over USD 70,000) are unaffordable. This paper describes the design, optimisation and pilot testing of a 7-degree of freedom (DOF) transhumeral motorised prosthesis controlled by TinyML-based myoelectric decoding. The gesture-recognition accuracy (F1-score) of 95.5% across five able-bodied participants (gesture-level accuracy: approximately 92.3% for the most difficult gesture) with an end-to-end response time of 100ms is achieved using an affordable ESP32 microcontroller and a Cost-Complexity Pruned (CCP) Random Forest (RF) classifier. This paper outlines the hardware-software co-design, mathematical formulation of surface Electromyography (sEMG) processing (digital Z-transforms) and a 60-second subject-specific incremental calibration routine. The design includes a dual-rail power supply network and an 8-channel analog multiplexer to address hardware limitations of low-cost microcontrollers. With a total bill of materials (BOM) of PKR 60,000-80,000 (USD 215-286) these preliminary findings are a good step toward value-based, low-cost assistive technology that may be deployed at low cost in resource-poor settings.

Submission Date: 15 Apr, 2026 Reviews Completed: 8 May, 2026
Acceptance Date: 16 May, 2026 Publication Date: 23 Jun, 2026