Hardware components | ||||||
 |
| × | 1 | |||
Software apps and online services | ||||||
 |
| |||||
Hand tools and fabrication machines | ||||||
| ||||||
We tried to deploy an impulse that is able to tell the difference between smoke, smoke from grilled meat and no smoke at all..
https://studio.edgeimpulse.com/public/31378/latest
Unfortunately it was designed for a sensor that outputs CO2 and TVOCs while the BME688 only outputs gas resistance, at least on the nrf Edge Impulse app, so we found out it wasn't compatible for inferencing with the Thingy:53.
Subsquently we decided to do our own data acquisition to build our ML model through Edge Impulse. It was during that stage that while experimenting with the Thingy:53 and its different firmwares namely the nrf Machine Learning we couldn't reflash it anymore after, neither through Bluetooth nor USB, and lacking a 10 pin SWD 1.27 mm cable we haven't been able to unbrick it yet despite our best efforts using mcumgr, etc..
And that was our path to try and get to know when the fire pit was being used without a line of sight.
Comments