Skip to the content
2 layersSTOF sensor, the printed circuit board in 3D

Low cost 16 IMU data glove with real time hand tracking and fingertip force sensing. BEng final year project

Copper layers
2
Thickness
1.6 mm
Parts
8
Nets
7
Pads
18
Tracks, vias
0, 0
Routed
Not yet
KiCad
KiCad 9.0 (file 20241229)
Created
18 May 2026
Updated
4 Jul 2026

Overview

STOF sensor is an open-source printed circuit board by OsamaAlsahli, from the OsamaAlsahli/data-glove repository on GitHub. It is a 2-layer board with 8 parts on 7 nets.

It is built around the TEMD7000X01. Its parts include 5 diodes, 2 connectors and 1 resistor. It belongs with the sensors and measurement and robots and motor control boards and the ESP32 boards of this library.

Main components

1

README

README.md

Data Glove

Low-cost, high-accuracy data glove with real-time hand motion tracking and fingertip force sensing.

Final Year Project — Osama Alsahli, BEng Electronic Engineering, University of York, 2026.

Data glove prototype with fingertip STOF sensor and IMU wiring

What this is

A wearable glove that captures finger and hand orientation using sixteen BNO055 9-axis IMUs (one per finger phalanx plus one on the dorsum), with soft optical tactile sensors (STOF) at the fingertips for contact force. Quaternion data is streamed from an ESP32 over USB serial to a Python bridge, which serves it to the FSGlove visualisation framework via gRPC.

Headline hardware bill of materials: 16× BNO055 IMUs, 2× TCA9548A I2C multiplexers, one ESP32 dev board, soft Ecoflex 00-30 STOF sensors at the fingertips. Total cost of purchased electronics ≈ £105.

For the full report (motivation, lit review, design, evaluation), see the accompanying BEng final report.

Repository layout

text
firmware/         ESP32 Arduino sketches (C++)
  adafruit_test/  Single-BNO055 smoke test
  tst/            Single sensor through TCA9548A
  tst2/           TCA9548A I2C scanner
  webserial_3d/   Single sensor for the Adafruit WebSerial 3D viewer
  quat_finger/    4-IMU single-finger demo
  quat/           Full 16-IMU glove firmware
  fsglove-main/   FSGlove visualisation framework (third-party, included
                  for reproducibility - see Licence section)

bridge/           Python serial-to-gRPC bridge
  bridge.py            Full 16-sensor bridge (the one you actually run)
  bridge_finger.py     Incomplete 4-sensor variant, kept for reference
  read_imu.py          Minimal serial quaternion printer
  visualise_imu.py     Live OpenGL "bone" visualiser

images/           Prototype photos, kinematic model, Gantt chart
schematics/       PCB / schematic designs (KiCad) for STOF and main board

How to flash the firmware

  1. Open firmware/quat/quat.ino in the Arduino IDE.
  2. Install the Adafruit BNO055 and Adafruit Unified Sensor libraries from the Library Manager.
  3. Select the ESP32 board, the matching USB serial port, and upload.

The single-sensor bring-up sketches (adafruit_test, tst, tst2, webserial_3d) are useful for verifying hardware before flashing the full build.

How to run the bridge

bash
cd bridge
pip install pyserial grpcio
# Generate imu_packet_pb2.py / imu_packet_pb2_grpc.py from the FSGlove proto
# (imu_packet.proto lives inside firmware/fsglove-main).
python bridge.py

The bridge listens on localhost:18890; start FSGlove afterwards and it will subscribe to the stream.

If you want to drive the visualiser without hardware, swap the target=read_serial line in bridge.py's main() for target=generate_fake_serial.

Licence

The code I wrote (everything under firmware/ except fsglove-main/, and everything under bridge/) is released for academic and personal use.

firmware/fsglove-main/ is included for reproducibility only and is the work of its original authors — see its own README and licence inside that folder or (https://github.com/davidliyutong/fsglove). The SMPLH model weights (SMPLH_male.pkl, SMPLH_female.pkl) required by FSGlove are not included in this repo. To obtain them and get FSGlove running, follow the setup instructions in the upstream FSGlove repository (https://github.com/davidliyutong/fsglove); the project page is at https://sites.google.com/view/fsglove/.

Known limitations

  • Single multiplexer firmware. firmware/quat/quat.ino was written assuming one 16 channel multiplexer, but the final hardware splits the 16 IMUs across two TCA9548As (each with 8 channels). Reading sensors 8 to 15 therefore needs a second multiplexer select layer that this build does not yet implement. The limitation is flagged inline with a NOTE: comment. The single finger and 8 sensor paths are fully working.
  • bridge/bridge_finger.py is an unfinished early version of the bridge, kept for historical reference. The working bridge is bridge/bridge.py.

Parts

8
QtyValueFootprintReferences
4TEMD7000X01LED_SMD:LED_0805_2012MetricD1, D2, D3, D4
2Conn_01x03Connector_Wire:SolderWire-0.1sqmm_1x03_P3.6mm_D0.4mm_OD1mmJ1, J2
1100Resistor_SMD:R_0603_1608MetricR1
1LEDLED_SMD:LED_0805_2012MetricLED1

More from OsamaAlsahli/data-glove

3
  • 2 layersSTOF sensor, the printed circuit board in 3D

    STOF sensor

    OsamaAlsahli/data-glove

    Low cost 16 IMU data glove with real time hand tracking and fingertip force sensing. BEng final year project

    7 parts
  • 2 layers10 × 149.9 mmSTOF v3, the printed circuit board in 3D

    STOF v3

    OsamaAlsahli/data-glove

    Low cost 16 IMU data glove with real time hand tracking and fingertip force sensing. BEng final year project

    7 parts
  • 2 layersSchematics, the printed circuit board in 3D

    Schematics

    OsamaAlsahli/data-glove

    Low cost 16 IMU data glove with real time hand tracking and fingertip force sensing. BEng final year project

    5 parts

Similar boards

  • 4 layers205 × 140 mmSpacescape, the printed circuit board in 3D

    Spacescape

    Napkin-to-Billboard/100GI

    100 Gadget Ideas - Napkin to Billboard

    359 partsMIT
  • 2 layers20.5 × 24.8 mmPi interface, the printed circuit board in 3D

    Pi interface

    opulo-inc/lumenpnp

    The LumenPnP is an open source pick and place machine.

    8 parts3.7k
  • 2 layers59.9 × 25.7 mmBuffer, the printed circuit board in 3D

    Buffer

    hoglet67/RGBtoHDMI

    Bare-metal Raspberry Pi project that provides pixel-perfect sampling of Retro Computer RGB/YUV video and conversion to HDMI

    8 parts1kGPL-3.0
  • 2 layers52 × 12.6 mmUSB, the printed circuit board in 3D

    USB

    shanteacontrols/OpenDeck

    A platform for building interactive control and sensing devices using MIDI and OSC built with Zephyr RTOS.

    8 parts960Apache-2.0
  • 2 layers87.5 × 48.5 mm7 input macropad, the printed circuit board in 3D

    7 input macropad

    hackclub/OnBoard

    💡 Join 1,000 teenagers and make your first PCB with a $100 grant!

    8 parts915
  • 2 layers36.2 × 63.1 mmPicoMemcard pico, the printed circuit board in 3D

    PicoMemcard pico

    dangiu/PicoMemcard

    Emulating PSX Memory Card (or controller) using a Raspberry Pi Pico

    8 parts694GPL-3.0

Questions

What parts does the STOF sensor use?

8 parts, built around the TEMD7000X01. The full list of its parts is on this page, with links to each part's symbol and footprint where PCBKit's library has them.

Where can I download the STOF sensor KiCad files?

From the OsamaAlsahli/data-glove repository on GitHub, at the commit this page was made from: the board is schematics/STOF_v1/STOF_sensor.kicad_pcb and its schematic schematics/STOF_v1/STOF_sensor-backups/STOF_sensor-2026-04-01_190337/STOF_sensor.kicad_sch. PCBKit opens the board in the browser, with nothing to install.

Can I use the STOF sensor design in my own project?

Its repository states no licence, so its author keeps all rights: ask OsamaAlsahli before reusing the design.