Over 20 Independent Buttons to Single Raspberry Pi with timer

Question

Okey, so I never had anything to do with a Raspberry Pi or Arduino and my question may sound dumb. I know some Python but I'm bad in electronics.

Lets say I need a system, where there are 10 work stations at an assembly line.

Now, the assembly lines speed is dependent of the slowest work station (it waits for all 10 work stations to finish). The system's job is to measure the efficiency of work stations. Simply put. At each work station, there should be a button and when the line starts to roll, a timer starts, and the button records the time for that particular work station. When each button is pressed, the timer resets, starts again and the cycle goes on. I know how to program that, I know how to get the data from Python code. I just have no idea how to make the hardware.

What would be better. Arduino or Raspberry Pi? Is there a way to get 20-30 independent buttons, which would be about 10 meters apart from each other? Of course, there should be some error handling. Like, someone may press the button but didn't intend so. So if someone is ready, he presses the button and changes the state of an variable and a LED lights up if he is ready. If he pressed it by accident, he just presses it again, or he has to press two buttons instead of one, so that it's safer against human error.

My main question to this would be, is this something hard to do? I'm a second year Computer Science student but we didn't cover a lot in Hardware sadly. Is this doable on one Raspberry Pi? If not, is it hard to connect a couple of Raspberries to have more USB/GPIO inputs? Or maybe should there be a microcontroller/chip?

I'm eager to learn and this, when going over it in my head, does not look like something very hard to do. Maybe some book recommendations or projects I should begin with?

Answer 1

The Rasperry Pi has lots of digital inputs (see here) so you can easily do as many buttons as digital inputs. Some models have 26 GPIOs, so that's 26 buttons easily.

If that's not enough, you can create a matrix. This is a little bit more complex, but really not very. There's a good explanation here.

In short, imagine you needed 100 buttons. You'd arrange them in a grid of 10x10. Each button would be on the junction between a row and a column: when you press it, it connects the row to the column. Each row and each column is connected to a separate GPIO - so you need 20 GPIOs in total.

Set your rows (say) to outputs and your columns to inputs. Enable the internal pull-up resistor for the columns - which means if nothing is connected to them, they'll read "high" (value 1).

Now to read your keys, you start at the first row: set its pin to low (=value 0, or 0v), and read each of the 10 columns. If any of the buttons on the first row are pressed, it'll pull the input down to 0v and you'll see 0 on the appropriate input, otherwise you'll see 1. Set it back to high (=value 1, or 5v) and repeat for all 10 rows, and you've read all 100 keys.

This approach needs no external components to understand or get working. The main downside is that you get ghosting of keys: if I press three or more keys together, depending on which combination it is, it can be hard to work out which have been pressed. For your scenario I'd have thought this was an acceptable compromise.

The other disadvantage is that you're going to have a huge jungle of wires, but you're hardly going to be able to avoid that to be honest.

This is how just about all keyboards and keypads work, by the way.