Final-Year Research Project · 2024 / 2025

Restore hand function with intelligent IoT rehabilitation, at home.

FLEXA is a comprehensive IoT-based platform for paralyzed hand recovery. It mirrors healthy-hand motion onto the impaired hand, delivers cloud-prescribed exercises, tracks grip-strength recovery with machine learning, and adapts therapy speed in real time based on patient emotion and physiology.

Finger-wise motion mirroring Cloud-prescribed therapy ML recovery prediction Emotion-adaptive control
FIG.01 · System overview FLEX · 5ch ESP32 · BRIDGE FIREBASE · RTDB ML · GRIP STAGE
<3°
Mean Angular Error
95.1%
Recovery Stage Accuracy
<200ms
Cloud Sync Latency
~33.5K
Hardware Cost (LKR)
Why it matters

Smarter rehabilitation for the post-stroke recovery gap.

Most patients receive supervised therapy only 2–3 times per week, far below the daily repetition that drives neuroplastic recovery. FLEXA fills this gap with a unified IoT platform built for the home.

24 / 7

Continuous monitoring

Real-time capture of finger angles, grip force, heart rate and patient emotion across every session, streamed to clinicians through Firebase.

Per-finger

Independent actuation

Five vacuum-driven soft actuators move each finger independently — a qualitative advance over uniform-actuation gloves on the market today.

ML-driven

Recovery prediction

A six-sensor grip device feeds a TensorFlow classifier that places each patient on a five-stage recovery scale with 95% accuracy.

Adaptive

Emotion-aware therapy

NeuroSpeed fuses ensemble facial-expression recognition with EAR, PERCLOS, head pose and HR to throttle therapy speed in under 250 ms.

Built for real users

Clinically supervised. Patient-friendly. Affordable.

Whether you are a physiotherapist managing a caseload, a stroke survivor doing daily home therapy, or a researcher exploring tele-rehabilitation, FLEXA delivers strong clinical oversight without enterprise hardware costs.

See how it works

Take the next step toward accessible hand rehabilitation.

Read the technical paper, review the documents, or get in touch with the project team for collaboration and evaluation.