We were too early regarding the technology (doing 3d real time reconstruction of inner rooms using laser scanner - 2011), without even owning a laser scanner (price tag: $30000). We got to rent one for some hours ($2000/4hours) to get some samples but they were too few to draw a generic, working algorithm. With the samples it worked well, but we didn't get the opportunity to move further, then we ran out of money (funded by a University Program, from the money we bought computers, office stuff and travels) pretty soon, because none looked pretty interested in the product. 3 years later, other technologies reached the market and it was 'successful', even though I don't see any disruption until today.

Learned a lot in laser scanning and improved my programming skills tremendously (was CTO and main IT guy after finishing university).

Unfortunately PCs were not so fast as today and there was no help like today with all those APIs available, but only 2 major libraries that helped me to achieve the MVP: OpenCV and PointcloudLibrary.