FlexiTac: An Open-Source, Scalable Tactile Solution for Robotic Systems

Authors:
Binghao Huang¹, Yunzhu Li¹
¹Columbia University

FlexiTac is an open-source, scalable tactile sensing platform designed to make touch sensing more accessible for robotics. It supports flexible sensor fabrication, real-time tactile readout, and integration into manipulation systems such as grippers, robot arms, and tactile skins. The broader FlexiTac project also includes hardware tutorials, simulation support, and system design examples for robotic applications. Project website: https://flexitac.github.io/

Overview

This repository contains the hardware-side code and examples for reading FlexiTac tactile sensors, including:

• Arduino firmware for tactile data acquisition
• Python visualization for real-time tactile readout

To build the tactile sensors, follow the instructions below: [Hardware Assembly Tutorial] .

Hardware Description

A FlexiTac sensor is scanned row-by-row and column-by-column by the microcontroller, then streamed to the host computer over serial.

In the current fast example:

• tactile resolution: 16 × 32
• serial baud rate: 2,000,000
• frame format: 2-byte header + 512 data bytes
• frame header: 0xAA 0x55
• data type: 8-bit unsigned values

The provided Python script initializes a baseline from the first few frames, subtracts the baseline, applies thresholding and normalization, and displays the tactile response as a heatmap.

Option 1 (Quick Start)

The fastest way to get running is to install the official flexitac package from PyPI. It bundles the firmware flasher, serial reader, and live heatmap as CLI tools — no need to clone this repo.

1. Upload the Arduino firmware

Open 32_16_fast.ino in the Arduino IDE, select the correct board and serial port, and upload it to the microcontroller.

After uploading, the board should appear as a serial device such as:

/dev/ttyUSB0

2. Install

pip install "flexitac[examples]"

The [examples] extra pulls in matplotlib for the live heatmap.

3. Stream / visualize

sudo chmod 777 /dev/ttyUSB*
flexitac-heatmap --port /dev/ttyUSB0      # live heatmap
flexitac-stream  --port /dev/ttyUSB0      # raw stream + FPS/stats

4. Use from Python

from flexitac import FlexiTacSensor

with FlexiTacSensor("/dev/ttyUSB0") as sensor:
    for frame in sensor:
        print(frame.normalized.shape, frame.normalized.max())

Source and issue tracker for the Python package: https://github.com/WT-MM/PyFlexiTac

---

Option 2 (Development Installation)

Use this path if you want to modify the firmware, hack on the visualization script, or develop against the raw hardware protocol. This is the source that the PyPI package wraps.

1. Upload the Arduino firmware

Open 32_16_fast.ino in the Arduino IDE, select the correct board and serial port, and upload it to the microcontroller.

After uploading, the board should appear as a serial device such as:

/dev/ttyUSB0

2. Install Python dependencies

Install the required Python packages:

pip install numpy pyserial opencv-python scipy

3. Check the serial port

Verify the device path on your machine:

ls /dev/ttyUSB*

If needed, update the port in the Python script:

PORT = "/dev/ttyUSB0"

4. Linux serial permission if needed

If you do not have permission to access the serial device, run:

sudo chmod 777 /dev/ttyUSB*

5. Run the example

Start the Python visualizer with:

python3 fast_32_16.py

🏷️ License

FlexiTac © 2026 by Columbia University is licensed under CC BY-NC 4.0. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc/4.0/

Acknowledgment

We thank Yiyue Luo, Devin Murphy, Michael Foshey, Xinyue Zhu, Yunxi Zhu, Naian Tao, Wesley Maa, Jimmy Wang, Binglin Wang, Irving Fang, Hongyu Li, Yixuan Wang, Hanxiao Jiang, Kaifeng Zhang, Changyi Lin, Xuhui Kang, Yuhao Zhou, Pokuang Zhou, Siyu Ma, Chang Yu, Yuqi Liang, Haonan Chen, Rao Fu, Jinzhou Li, Yuming Gu, Fan Cheng, Xinyi Yang, Mingtong Zhang, Baoyu Li, Toru Lin, Haozhi Qi, Iretiayo Akinola , Jie Xu, and Yu-Wei Chao for their valuable discussions and support.