Skip to content

Overview

An Affordable data acquisition system with cellular connectivity, providing low-latency personal performance data, in the palm of your hand.

Our CompSci Capstone team is partnering with an ECE team to develop a complete sensor data acquisition device. With the ECE team's hardware sensors providing us data, we're sending data with an LTE-capable platform to an online backend. This is all accessible from our online, web-based client.

Project Core

The Problem

  • Current solutions are expensive, at $5,000+.
  • There are often many individual parts.
  • Too the average layman, interfaces are too complex.
  • A steep learning curve is often required to use current solutions effectively.
  • Comparison with existing solutions can be found in the requirements document.

Target Users

  • OSU Global Formula Racing Team
  • Small labs, startups, educators, hobbyists
  • Go-Karts, Snowmobiles, Mountain Bikes

Our Solution

  • Bridges the professional-hobbyist gap
  • Cost effective, around $500
  • Single Purchase Product (Pi w/ sensors)
  • User friendly interface for non-experts
  • Key technologies can be found in the design document.

The Communications Module

  • The Raspberry Pi 5 provides a high-performance Linux environment to run Rust programs similar to a full PC while also allowing an LTE module to be mounted via a HAT.
  • A power delivery circuit allows most off-the-shelf USB-C power sources to provide enough current for the Pi.
  • The data is gathered from a collection of sensors containing GPS, force, string, linear, and 9 degrees of freedom sensor.
  • The sensors record and transmit at 100 hz for high precision which produces 5.8 KB/second with future compression possibilities.
  • With constant 24 hour use the sensors produce about 348 KB/min, 20.9 MB/hour, 501 MB/day, 3.5 GB/week, or 15 GB/month.

Technical Highlights

  • Pi receives ECE Team’s data via USB or WIFI
  • Pi transmits data batches to AWS via 4G LTE
  • Pi transmits low-latency data via WebSockets
  • Flexible SQLite DB can allow new sensor types
  • Scaliable TCP server utilizing pattern matching
  • Web-based data visualizations
  • All software is written entirely in Rust

Diagrams

Data Acquisition System Overview

Database Entity Relationship Diagram

Architecture Design Diagram