pic: Roborio mounting and protective case

[Andrew Schreiber]

Originally Posted by faust1706

This is a bit off topic, but I had an interview today for an internship for developing software. One of the questions was how I unit test (stress test) my code. It is important to do, and it could be the difference between getting a job or not. You never know.

And if you really don't want to do it, I'm sure there exists some program online that allows you to upload a cad file (I don't even know what extension they have....) and it will do it for you, or you could figure out how to do it in whatever software you designed it in.

I could easily see the quality award going to a team that has documentation about their robot's durability (and of course a good robot) How much the frame will deflect during a collision, stuff like that. It shows good engineering practice.

Unit Testing is GREAT. But it's not the end of your tests. You need to do full integration testing as well as user acceptance testing. Furthermore, writing tests is an art in of itself. I've seen lots of situations where overconfidence in our test suite let a bug slip in. Or where the tests, while exercising the code in the proper ways, weren't quite testing every data point. (This was a tool for generating QRDA CAT 1 documents from a data warehouse for patients, testing every single data element wasn't required, just the ones we needed at the time). You've also got to worry about ACTUALLY stress testing your system as well as doing security audits and tests.



I can safely say that there is NOT a tool like that. FEA is a topic entire courses are dedicated to because, not only is it hard, it's complicated to set up your simulations. Even IF there were a tool like you described I feel it would cause more harm than good because of aforementioned terrible simulations. This is a situation where your ego is going to take a bit of a bruising - the CS world is easy compared to the ME world. I say that as a programmer. We control inputs, can filter bad ones, and testing all our edge cases is fast/easy. The real world? Not so much. Testing a design is often much more difficult than testing some code.



I highly suggest reading the criteria for the Quality Award in the manual. I'd also suggest reading the sections on Industrial Design Award. While your gut feeling is mostly correct there's a fair bit more that goes into deciding those awards than what team can show the most pretty pictures of contrived scenarios.