Strate App

HW and SW solution to detect enduro bike crashes | BSc project | 2019/2020 | Supervisor

The initial idea was proposed by Pedro Cruz, a BSc student which practices Enduro motorcycle sport – create a simple system to detect falls during his Enduro trips and message his emergency contact.

Architecture of the prototype with the various services communicating through the message broker.

Although being a bachelor degree project, it started earlier (in September 2019) and exposed the student to a myriad of different concepts:

  • Hardware / device creation (3D printed box, ESP32, MPU6050, battery)
  • Real-time operating systems programming (FreeRTOS) with parallel tasks accross the available cores
  • Mobile app development (React Native Android app that controls the hardware and reads sensors data via Bluetooth)
  • Backend development (using Node.JS and MongoDB)
  • Introduction to pattern recognition (data analysis, simple decision trees, possibility of exploring simple machine learning concepts)
On the left, the proposed hardware solution. Right, the android app logging data from the current trip.

The year long project was divided in several mini-projects, from hardware testing, selection and design. Developing the freeRTOS code, Android App, collecting some example datasets, API+frontend and data analysis.

The HW part was one of the major challenges, testing different prototypes and even printing a 3D case.

Evolution of the HW prototype, from the first version (tupperware) to the last (printed 3D case).

Although unfinished, the last part consisted in analysing data from trips and simulated crashes.

Sensor data analysis and visualization for a short test trip.