Sensor-fusion know-how: I can make noisy sensors tell the truth
Posted by Pedro · Location-independent
A skill nobody pays me for yet: sensor fusion. I can take messy readings from cheap sensors (accelerometers, gyros, GPS, magnetometers, airspeed probes) and combine them with Kalman filtering into a position, heading or speed estimate you can trust. It's the same maths that keeps drones level and phones knowing which way they face. My background is embedded systems, RF hardware and signal processing, and I'm an engineer in Casablanca.
A skill nobody pays me for yet: sensor fusion.…
Why I’m posting this
This maths stays locked in aerospace firms. I'd love to see it help a farm, a sport, a school or an art piece.
Open to
Collaboration, Ideas
Free public Ideas
I love this post!! Always thought our phones measure so much more than we ever use! Some crazy uses there for sure, maybe partner with some app developers?
For livestock, could have some collars that will give the data to the owner on their animals (if they are not moving, limping, laying down a lot?) - can be very useful info?
I am totally not good at physics but can this be used in mapping contaminated land from the air? If so - demining (I am in Ukraine) is a very well funded and needed field!
A bicycle cargo-trailer maker could be an interesting smaller-scale partner: compare how different loads affect movement on the same route. They bring the physical prototype and test rides; you help turn noisy measurements into a useful comparison. Is analysing recorded data a practical way to begin remotely?
Ideas are free public suggestions. Proposals are paid and private and are never shown here.