BIONIX: A Wireless, Low-Cost Prosthetic Arm with Dual-Signal EEG and EMG Control

Affordable upper-limb prostheses often lack intuitive control systems, limiting functionality and accessibility for amputees in low-resource settings. This project presents a low-cost, dual-mode neuro-muscular control system integrating electroencephalography (EEG) and electromyography (EMG) to enable real-time, multi-degree-of-freedom control of a prosthetic arm. EEG signals are acquired using the NeuroSky MindWave Mobile 2 and transmitted via ThinkGear Bluetooth packets to an ESP32 microcontroller running a lightweight classification model. The model was trained on 1500 seconds of recorded EEG data using a 6-frame sliding window with low-pass filtering, excluding poor-signal samples and using a 70/20/10 training–validation–test split. The classifier detects strong blink events, which toggle the hand between open and closed states. EMG signals are acquired using a MyoWare 2.0 sensor and SparkFun wireless shield and transmitted to a second ESP32, which performs threshold-based detection. Three activation bands (rest: 0–T1; extension: T1–T2; contraction: greater than T2) enable intuitive elbow control, with movement triggered only after eight consecutive frames in a movement class to improve stability. The EEG-controlled ESP32 actuates four finger servos, while the EMG-controlled ESP32 drives two elbow servos. A functional prototype was constructed using low-cost materials (total cost approximately 240 dollars), with most expense attributed to the commercial EEG headset. Future work includes transitioning to a 3D-printed chassis, integrating auto-regressive models to reduce EMG latency, and upgrading servo torque for improved load capacity and grip strength. This system demonstrates a feasible pathway to low-cost, biologically intuitive prosthetic control suitable for underserved and global health applications.

💡 Research Summary

This paper presents a low‑cost, wireless upper‑limb prosthetic arm that combines electroencephalography (EEG) and electromyography (EMG) to control both hand grasp and elbow flexion/extension. The EEG subsystem uses a NeuroSky MindWave Mobile 2 consumer headset, which streams ThinkGear Bluetooth packets to an ESP32 microcontroller. Raw EEG samples (or band‑power metrics) are filtered with a first‑order low‑pass filter, then framed using a six‑frame sliding window (≈300 ms total) with a hop of one frame. A compact CNN‑LSTM network, trained on 1 500 seconds of labeled data (blink vs non‑blink) with a 70/20/10 split, performs blink detection. The model is quantized to 8‑bit TensorFlow Lite Micro for on‑device inference, achieving real‑time latency on the ESP32. When a strong blink is detected, the system toggles the hand between open and closed states by driving four finger servos.

The EMG subsystem employs a MyoWare 2.0 sensor and a SparkFun wireless shield. The shield transmits a rectified envelope of the muscle activity to a second ESP32. After low‑pass filtering (≈20–30 Hz cutoff), the signal is compared against two thresholds (T1, T2) that define three activation bands: rest (0‑T1), extension (T1‑T2), and contraction (>T2). To improve robustness, a command is issued only after eight consecutive frames fall in the same band, which reduces false triggers caused by noise or brief artefacts. This ESP32 drives two higher‑torque servos that produce elbow flexion and extension.

The hardware architecture is modular: one ESP32 handles hand control, the other handles elbow control, and both communicate wirelessly with the sensors. The prototype is built from inexpensive components—two ESP32 boards, a NeuroSky headset ($100), a MyoWare sensor and shield ($60), servos, a 5 V LiPo battery, and a temporary structure made of plywood and cardboard. The total bill‑of‑materials is approximately $240, with the EEG headset accounting for the largest share.

The authors discuss related work, noting that most low‑cost prostheses rely solely on EMG (limited to simple grasp) or on EEG (often single‑degree‑of‑freedom hand control). Hybrid approaches have been demonstrated in laboratory settings but rarely with consumer‑grade, wireless, and affordable hardware. This system fills that gap by providing a dual‑modal, multi‑joint control platform that can be assembled with readily available parts.

Limitations are acknowledged: consumer EEG suffers from low signal‑to‑noise ratio and limited channel count, making blink‑based hand control susceptible to fatigue and mis‑classification; EMG performance degrades when residual muscles are weak; the servos used have modest torque, restricting load‑bearing capability; and overall latency, while acceptable for basic tasks, could be reduced further. Future work includes redesigning the chassis with 3‑D‑printed PLA components, integrating auto‑regressive models to predict EMG and lower latency, upgrading to higher‑torque gear motors for stronger grip, and exploring more sophisticated deep‑learning architectures (e.g., transformers) to decode richer EEG patterns for wrist or multi‑joint gestures.

In summary, the paper demonstrates a feasible, low‑cost pathway to biologically intuitive prosthetic control by fusing EEG‑driven blink detection with EMG‑based elbow actuation. The approach is reproducible, modular, and targeted at underserved populations, offering a practical blueprint for expanding access to functional prosthetic technology in low‑resource settings.