In my last year of college, a couple friends and I ended up working on implementing a vehicle-to-infrastructure communications demo for the department of transportation. We were doing it for a grade in a special projects class, but we were working with a consulting company that was being paid by the DOT to implement the demo. Toward the beginning of the project, the consulting company folk were very concerned about giving college students any non-trivial amount of scope, and were talking about how they would hedge all their bets by implementing everything themselves and only use our stuff if it panned out.
The demo itself consisted of about a dozen different scenarios. The scenarios were all basically some form or another of geofencing, and it made sense to make a simple framework to get 90% of the way, then specialize for each scenario. The consulting company didn't see it that way, and instead wanted to treat each scenario as a separate unit of work.
Fast forward to the end of the semester, and my friends and I demoed our framework for the professor, and a Motorola radio rep. It all worked and we got A's. It was like 400 lines of python. A couple weeks before the DOT demo, we started seriously trying to integrate with the consulting company's stuff, and it was laughably bad.
The consulting company knew they dropped the ball, but figured the three of us could just scramble to finish it all on top of our framework. The Motorola rep chimed in and pointed out that we already got our A's, and that the consulting company was getting paid $500k. They ended up paying us something like $20K, and it only took us a few hours to implement all the scenarios on top of our framework. The demo went well, and we ended up directly helping the DOT demo it a few more times over that summer.
The demo itself consisted of about a dozen different scenarios. The scenarios were all basically some form or another of geofencing, and it made sense to make a simple framework to get 90% of the way, then specialize for each scenario. The consulting company didn't see it that way, and instead wanted to treat each scenario as a separate unit of work.
Fast forward to the end of the semester, and my friends and I demoed our framework for the professor, and a Motorola radio rep. It all worked and we got A's. It was like 400 lines of python. A couple weeks before the DOT demo, we started seriously trying to integrate with the consulting company's stuff, and it was laughably bad.
The consulting company knew they dropped the ball, but figured the three of us could just scramble to finish it all on top of our framework. The Motorola rep chimed in and pointed out that we already got our A's, and that the consulting company was getting paid $500k. They ended up paying us something like $20K, and it only took us a few hours to implement all the scenarios on top of our framework. The demo went well, and we ended up directly helping the DOT demo it a few more times over that summer.