• Question: how many times do you have to improve your prototypes

    Asked by whyyto to Dawn, James, Sarah, Sylvain, Tomas, Vaanu on 11 Nov 2016.
    • Photo: Sylvain Jamais

      Sylvain Jamais answered on 11 Nov 2016:

      There is no hard and fast rule, typically, the more difficult your objective, the more iterations you will need to get it right, but there are other ways than prototypes to do that such as calculations and simulations. the right combination of theory and experiment usually means you have to do a lot less prototyping, but you always end up building something for the final proof.

    • Photo: Dawn Gillies

      Dawn Gillies answered on 11 Nov 2016:

      It totally depends on the project, and how much preparation you have done. I have just made a new sample holder and I’ll be testing it next week, I’m really hoping it works!! Usually when I’m desiging something I try to go through a few designs before I actually make something to avoid as many problems with a protoype as I can think of. This usually reduces the amount of improving I have to do.

    • Photo: Vaanathi Sundaresan

      Vaanathi Sundaresan answered on 11 Nov 2016:

      That depends on the problem I an trying to solve and the complexity of the solution. Usually we develop a prototype on a specific data and test it and completely new data to check how well it is performing. If the variability of test data is high (or if the test data does not resemble training data much) then we need another revision to our prototypes. On an average, we need to develop at least 4 or 5 versions of prototypes

    • Photo: Sarah Hampson

      Sarah Hampson answered on 11 Nov 2016:

      Like the others have said, it depends on how complicated the thing you’re making is.

      1 of my microchips is on the 2nd prototype stage and it still isn’t working at all. But another is working- it’s on the 4th stage right now.

    • Photo: James Clarke

      James Clarke answered on 17 Nov 2016:

      Usually 2 or 3 times, but occasionally more. In fact I think there’s only one hard and fast rule for this: nothing works first time.

Comments