What is simulation? In laymen’s terms it is making a mathematical model of reality to try to predict how a system will behave under certain conditions. Before computers, this was all done on paper. With compute software, more and more simulation tools are developed and even visualization options are integrated so you can see how the scenario will look like and not just use numbers.?
No. As people discovered, there are uncertainties and deviations between simulation and reality. Due to that, they developed an incremental test approach. That is by incrementally adding element from the final product to the test. That is done on several platforms.
- On Computers and that is referred to as Model in the Loop (MIL) or Software in the Loop (SIL) where you use only software intended to go on the final product and test it with simulated inputs and determine if the software gives the intended output.
- Processer in the Loop (PIL) and that is when you have the “onboard computer” of the product, more commonly known as microprocessor, controller or control unit; flash the software on it and again, simulate with your inputs to observe what the outputs would be at the pin level. The pin in this case are similar to what you see inside the plug of your cable but on the electronic hardware.?
- Hardware in the Loop (HIL) and here you bring in more electronic and perhaps even mechanical arts to test the system with again with simulation
- Dynamometer/test cell: in this case you increase more parts and add external drivers that can give you loads and more realistic outputs.
- Prototype: now you are testing with the full product.
Why all this? Why not go to the prototype directly?
I have had experience working for a company that for a long time thought it was cheaper to produce prototypes than to invest in simulations and incremental testing. That was the way they always have done it. Here is why they shifted.?
- They did not have to wait to have problems. With simulations available on computers, you can make a lot of fast tries and test many scenarios very fast. The faster you find problem, the cheaper it is to fix them.
- The root cause of the problems were easier to determine. When you test things in smaller increments and at a smaller level, it is easier to isolate problems with that part or the interaction between these parts. The more integrated components you have, the harder it is to determine why a problem happens.?
- The scale of damage to the prototype initially. with the rising complexity of mechatronic systems, more problems are discovered and many caused damage to expensive hardware parts.?
- The simulation tools are configurable and reusable across programs.?
Overall, when you combine these arguments, the initial capital investment to setup simulation options was smaller compared to relying on prototypes for all the testing.
It is harder now to find a product developer that fully relies on prototypes now than does full simulation scope due to the power of simulation.
