Vendel

This week we have done more to get started on the project, doling out group roles and discussing what we are supposed to do towards the project individually and according to our specializations as computer, machine and electrical engineering students.

I took the role as team leader for the project, though I don’t have much experience with that, but this is a learning opportunity, and I try to take it very seriously!

Andreas offered to be responsible for posting the blog weekly as we finish it, arguing that Bendik and I, being the computer engineers on this group with a project that focuses heavily on the computer aspect of the trifecta, would likely be too swamped with work for it to be reasonable for us to be the ones to do that extra of work.

Marte seemed to be geared up for dealing with the Teams-group we made for file-sharing and video/voice chat purposes.

Sadly, as we lost Simone early, we also lost the member of the group who seemed the most proficient with Azure DevOps, and so it has felt, at least to me, more difficult to use as a result.

My focus as a computer engineer has been on the deeper brain of the mouse, how it should memorize the maze, and how it should traverse it. Having been busy with a tight schedule of Game Development this week, I have not been able to devote as much time to this pursuit as I wanted.

I have been able to determine that the mouse should see the maze as a 2D array, holding the statistics of the maze cells. These statistics should be whether or not there are walls on the four cardinal directions, and how far the mouse thinks the cell is from the goal.

For testing purposes, I also made a quick pseudocode of how I envision the mouse to traverse the maze, grossly simplified of course, as we still do not have a working prototype (at the time of writing).

So it will, hopefully, turn towards the lowest valued adjacent cell, and move there! Proof is in the pudding though, if it will do any of the things I expect it to do when actually testing.

Andreas

Week 2 has been filled with far more concrete work than week 1. This makes the end goal of an autonomous MicroMouse seem much more achievable than expected.

Firstly, a preliminary design for the IR depth sensor was made and will hopefully be tested with circuits when I receive the components from Bendik next week.

As well as designing the circuitry for the depth sensors, I have considered how we can use analog inputs of the Kitronik board to use all four sensors without interfering with each other.

After looking into the motor controller, it seems it does not support varying speeds for the motor. This is mainly because the controller outputs  to the motors instead of a controllable voltage. This will suffice as long as we input a 3V source to the controller. This will suffice for a minimally viable product, though for further development we could either use PWM or create a circuit to enable multiple speeds.

For the next week, I hope to fully complete the necessary circuit for the IR depth sensor and work with Marte to find a solution to the mounting of the circuitry and components so that we can have a physical prototype to perform tests upon.

Bendik

In week 2, I have been looking at and learning about how the IR depth sensor will work. It consists of an IR Infrared LED and a Photodiode Receiver. The way it works is the IR LED send out an infrared signal, which will then bounce off the wall and back into the IR receiver. The receiver will measure this signal and output an analog signal into the microbit. Depending on the strength of the signal (the analog value received), we can calculate about how close the IR depth sensor is to a wall.

I am hoping to be able to program and do a simple test of the IR depth sensor before our next meet. This way we can use the meet for next steps and start thinking out how the pairs should be placed and directed to ensure proper detection of the environment.

I was hoping to use more time on this this week, but game design sadly took more of the time this week, but that should be better going forward as this week was much more intense in that subject then the rest of the weeks will be.

Marte

Monday morning started with a short meeting with all the team members followed by a GitHub lecture. It was interesting learning about GitHub, but I don’t think I will use this being the mechanical engineer of this project.

After the lecture we received a «starter-kit» from Steven, containing different components that will make up some of the hardware of our micromouse.

We then proceeded to get to know the given components by retrieving data sheets. I continued to measure dimensions that were not given, and took some pictures of the components. I spent the rest of the day 3D-modelling the parts. I started modelling way too detailed, and found out I will do a more coarse modelling further considering this being a minimal viable product (MVP).

The goal for next week will be to continue with the product design and collaborate with Andreas to plan the MVP further.

Summary

As a conclution to the week we have made progress on researching how we can solve the most prevalent issues when designing a MicroMouse. This has not necessarily produced a lot of concrete documentation or graphs that can be shown in a pretty way. Despite this, it has been an important week for shaping the outlook for our MicroMouse.

“Perseverance is not a long race; it is many short races one after the other.”

– Walter Elliot