IoT Gait Analysis Insole using Pressure Sensors

Clinicians assess walking problems with force plates that cost lakhs and live in hospital labs — far out of reach of a student project, a physiotherapy student, or a sports coach. This project shrinks the concept into a shoe: eight force-sensitive resistors laminated into a flexible insole, sampled at 50 Hz by a small Bluetooth microcontroller, streamed to a web app that draws the foot's pressure map live. The app records walking sessions and computes the parameters gait analysis actually uses — cadence, stance/swing split, double-support time, peak pressures per zone and left-right symmetry — then plots the average gait cycle and flags asymmetries beyond a configurable threshold. The student gets a genuine embedded + signal-processing project: sensor interfacing, wireless streaming, real-time visualization and biomechanical parameter extraction. Because sensor placement, shoe fit and walking surface all affect readings, the kit makes no clinical claim — it is a screening and learning aid, and the report presents the student's own measured session data with a documented calibration procedure.

1. The sensor insole is placed in the student's shoe and paired to the web app over Bluetooth Low Energy.
2. The ESP32 samples all eight force sensors at 50 Hz, applies the calibration curve, and streams pressure packets to the browser.
3. The live page draws the plantar heatmap, updates per-zone kilopascal readouts, and plots the rolling left/right force curves.
4. The student performs a walk test (e.g. 10 metres); the app detects steps from the force signal and records the session.
5. The analysis page averages the steps into gait-cycle curves, computes symmetry bars and stride parameters, and renders the foot-strike assessment.
6. Session data exports as CSV; the report presents the student's own walk-test results with the calibration method and a clear non-medical disclaimer.

Tech stack:

• 8× FSR-402 thin-film force sensors on flexible insole substrate
• ESP32 BLE microcontroller, LiPo cell with charging module
• GaitLab web app (BLE WebBluetooth, canvas heatmap and charts, CSV export)
• 3D-printed ankle-mount enclosure for electronics (optional wear config)
• Calibration rig procedure (body-weight reference method)
• Session analysis engine (step detection, cycle averaging, symmetry metrics)

• [8-zone pressure sensing] Thin-film force sensors under heel, midfoot, metatarsals and toes capture the plantar pressure distribution at 50 Hz per foot.
• [Live pressure heatmap] The web app renders the foot's pressure map in real time over BLE — the demo centerpiece that makes the project instantly understandable.
• [Force curve recording] Left and right total-force traces are plotted live and stored per session, showing the characteristic double-bump of a normal step.
• [Gait-cycle analysis] The analysis page averages detected steps into heel/midfoot/forefoot pressure-vs-stance curves with loading, mid-stance and push-off bands marked.
• [Symmetry comparison] Peak pressures per zone are compared left vs right as paired bars, with a configurable asymmetry flag threshold for the report.
• [Stride parameters] Cadence, step count, stride length estimate, stance/swing percentage and double-support time are computed per session and exportable as CSV.
• [Calibration procedure] A documented body-weight calibration routine converts raw ADC readings to kilopascals, and the report explains the procedure and its limits honestly.

• Working sensor insole pair (8 FSR zones per foot, flex cabling, BLE module, battery)
• GaitLab web app source code (live + analysis pages, documented)
• Project report PDF (sensor interfacing, BLE streaming, gait parameters, buyer's walk-test results)
• PPT presentation for final review
• Viva Q&A preparation document (FSR principles, gait cycle biomechanics, BLE, signal processing)
• Calibration and walk-test protocol guide (step-by-step, with data sheet template)

• This is an educational screening aid, NOT a medical device: it does not diagnose any condition, and the app and report carry that disclaimer prominently.
• Pressure readings are indicative: FSRs drift with temperature and bending, so the calibration procedure must be followed and results presented with their tolerance.
• Only one insole is fully instrumented in the base kit; the second foot is mirrored in software unless the dual-insole option is ordered.
• WebBluetooth needs Chrome/Edge on desktop or Android; iPhones do not support it — the buyer needs a compatible device for the live demo.
• Battery life covers a demo session, not all-day wear; this is a lab instrument, not a consumer wearable.
• Results vary with shoe fit, socks, walking surface and speed — the protocol guide standardizes the walk test the student reports.