# TinyML Deployment Issue

**URL:** <https://discuss.tinyml.seas.harvard.edu/t/tinyml-deployment-issue/1655>\
**Category:** Projects\
**Created:** [December 19, 2024, 6:25am UTC](https://discuss.tinyml.seas.harvard.edu/t/tinyml-deployment-issue/1655 "2024-12-19T06:25:36Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![MuhammadWaqasAli](https://avatars.discourse-cdn.com/v4/letter/m/848f3c/32.png) [@MuhammadWaqasAli](https://discuss.tinyml.seas.harvard.edu/u/MuhammadWaqasAli)\
**Post date:** [December 19, 2024, 6:25am UTC](https://discuss.tinyml.seas.harvard.edu/t/tinyml-deployment-issue/1655/1 "2024-12-19T06:25:36Z")

</div>

Hi everyone,

I’m working on a project where I use TinyML to predict horse activity based on 6-axis accelerometer and gyroscope data. The data is collected via an ESP32-S3, and I’m trying to deploy a TensorFlow Lite (TFLite) model on an ESP32-S3 for real-time predictions.

However, I’m running into several library compatibility and deployment issues. Despite following standard TinyML deployment practices, the integration of the model with the ESP32-S3 seems problematic, especially when dealing with the TFLite Micro runtime. I’ve tried a few optimizations but still can’t get things running smoothly.

I’d appreciate any advice on:

1. How to effectively deploy a TinyML model on ESP32-S3.
2. Alternative approaches or tools that might work better for this use case.

Any tips, documentation, or experiences with similar setups would be a huge help!

Thanks in advance!