

This is a research project. We are planning to make an open source dataset of Bengali Sign Language (BdSL). To collect data of hand movements we are creating this smart glove for both hand using esp32, MPU650 and 3d printed flex sensors.
- Size
- 38.5 × 34.5 mm
- Copper layers
- 2
- Thickness
- 1.6 mm
- Parts
- 16
- Nets
- 34
- Pads
- 79
- Tracks, vias
- 107, 12
- Pours
- 1
- Routed
- Yes
- KiCad
- KiCad 10.0 (file 20260206)
- Created
- 10 Feb 2026
- Updated
- 31 Oct 2026
Overview
Glove is an open-source printed circuit board by Adnanosman9, from the Adnanosman9/Smart-Gloves repository on GitHub, published under the MIT licence. It is a 2-layer board of 38.5 × 34.5 mm with 16 parts on 34 nets, routed with 107 tracks and 12 vias.
It is built around the MPU-6050. Its parts include 7 resistors, 5 connectors, 2 capacitors and 2 ICs. It belongs with the ESP32 and RISC-V boards of this library.
Main components
1README
README.mdBdSL Translator Glove
The BdSL Translator Glove is a wearable device designed to translate BdSL into words without any camera setup.
How It Works
- Sensors: The glove tracks movement using bend sensors or custom, flexible 3D-printed tpu that changes its voltage when your fingers move.
- Brain: An esp module tracks exactly how your hand moves and tilts, turns those gestures into data, and sends the translated speech to a device.
Recent Design Updates
- Lower Cost: Switched to smaller sensors, cutting the hardware budget down to about $100 for a full two-handed setup.
- Better Fit: Upgraded the design from a clunky plastic box to form-fitting, flexible finger guides that follow the shape of the hand.
Finger Guides
Shoutout to zackfreedman for the original parametric ring glove concept that inspired this design. You can grab my modified STL here.
Instead of messing around with messy glues or tape, each finger uses a 3D-printed TPU channel segment where the sensor just slides right in and locks mechanically. I split these into two sizes:
- Long: For your index, middle, and ring fingers.
- Short: Form-fitted for the thumb and pinky to match their shorter lengths and natural joint angles.
Any standard FDM printing service can easily run these off in white TPU filament.
PCB Design
I put together a compact wrist hub in KiCad to keep the electronics tidy. A few design choices:
- Brain: The XIAO ESP32-C3 was a no-brainer. It's incredibly small, handles Bluetooth LE natively, and has just enough ADC pins for all 5 fingers.
- Hardware Filtering: Because flex sensors have a pretty high impedance, I added a 47kΩ voltage divider on every sensor input to keep readings steady. I also threw in 0.1µF decoupling capacitors along the sensor traces to smooth out some electrical noise I ran into during early testing.
The KiCad DRC checked out cleanly with 0 errors (just a few minor silkscreen overlap warnings that won't disrupt manufacturing).
Schematic:

PCB Layout:

Enclosure
The wrist housing was designed in Fusion 360 to keep the pcb safe.
Get the CAD assembly here
Alternative of commercial Sensors
Here is an alter for commercial sensors for cost optimization.
Firmware Status
Just a quick heads up: I haven't been able to run this code on physical hardware yet. It's highly likely there are some I2C pin configuration mismatches, flipped GPIO definitions, or sensor initialization bugs waiting to happen on the first boot. Please double-check your schematic assignments against your software configuration before flashing!
Bill of Materials
You can view the full parts list in the BOM file here.
GitHub @Adnanosman
Parts
16| Qty | Value | Footprint | References |
|---|---|---|---|
| 5 | 47k | Resistor_SMD:R_0805_2012Metric | R1, R2, R3, R4, R5 |
| 5 | Conn_01x02_Pin | Connector_PinHeader_2.54mm:PinHeader_1x02_P2.54mm_Horizontal | J1, J2, J3, J4, J5 |
| 2 | 0.1uf | Capacitor_SMD:C_0805_2012Metric | C1, C2 |
| 2 | 4.7k | Resistor_SMD:R_0805_2012Metric | R6, R7 |
| 1 | MPU-6050 | Sensor_Motion:InvenSense_QFN-24_4x4mm_P0.5mm | U1 |
| 1 | XIAO-ESP32-C3-SMD | RF_Module:MCU_Seeed_ESP32C3 | U2 |
Similar boards
- 2 layers23 × 120 mm

- 4 layers95 × 50 mm

Control board
ZhuYanzhen1/MM32_CLion
本项目是全国大学生智能车竞赛GDUT 2022年极速越野组的项目文件。采用了LQR控制器和多速率卡尔曼融合算法,并在单片机上实现。This project is the project document for the GDUT 2022 Extreme Off-Road Category of the National Student Intelligent Vehicle Competition.
129 parts58AGPL-3.0 - 2 layers50 × 66 mm

- 2 layers36 × 43 mm

Asac fc rev a
victorhook/asac-fc
asac fc is an rp2040 based flight controller made for FPV drone flying
119 parts52GPL-3.0 - 4 layers305 × 208.3 mm

- 2 layers98.5 × 17.8 mm

- 2 layers36.2 × 36.2 mm

- 2 layers26 × 21.3 mm

PModBoard
pms67/PModlet
PMOD PCB with USB-to-UART converter and IMU (MPU-6050, I2C) for FPGAs.
19 parts15MIT
Questions
How big is the Glove PCB?
It measures 38.5 × 34.5 mm and has 2 copper layers, 1.6 mm thick.
What parts does the Glove use?
16 parts, built around the MPU-6050. 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 Glove KiCad files?
From the Adnanosman9/Smart-Gloves repository on GitHub, at the commit this page was made from: the board is PCB/glove.kicad_pcb and its schematic PCB/glove.kicad_sch. PCBKit opens the board in the browser, with nothing to install.
Can I use the Glove design in my own project?
Its repository publishes it under the MIT licence: you may reuse it on that licence's terms (keep the author's notice, and for a copyleft licence share your changes alike).