This topic contains 3 replies, has 1 voice, and was last updated by 
-
Posts
-
My first day was very routine; it was lots of paperwork and security briefs so I was expecting all that. I met my mentor, Jon, and he sat me down to explain my main project for the four weeks. Basically it was completely not what I was expecting, involving lots of coding and programming and not much of what I do at school…So it was a bit of a shaky start, as he slowly realised that I knew less and less of what he was expecting! When I got home from my first day I was quite worried about my 4 weeks, and whether I would enjoy them. My second day however – much better. The night before I had stayed up late looking at books and articles on how to use human readable code, and getting familiar with the Arduino integrated development environment (what I was using as a platform for my coding). This will sound quite nerdy, but I’m getting very into the electronics and software behind my project and it’s sort of becoming a hobby at home; I’ve bought my own Arduino Uno board for twenty pounds and a load of electrical components complete with breadboard for another tenner. It sounds quite sad sitting up in my room playing with all this stuff, but I listen to music while I work so it’s quite therapeutic.
At the moment I’m working on figuring at an algorithm for detecting explosions underwater. The setup is quite simple, 8 sensors around the circumference of a plate, that will each listen for sound waves passing them. The difficult bit is figuring out an algorithm that uses the times that the sensors go off to calculate a bearing of where the explosion is coming from. Once I’ve done that I’ll be working on putting that into a program for my lovely Arduino board to do all the hard work instead.
What I’ve really liked about this so far is that the work I’m doing is actually going to be achieving something, and I’ll be able to see how well it works on the AUV that it will be used on. I’m in an office with two engineers, who are working on programming the software for a generator for an unmanned boat. It’s so cool to see them work, because they keep facing lots of different problems, many more than I thought engineers actually did, and then both work on trying different solutions until they fix it. They actually get really hyped when they sort it, because it will usually be a really clever fix that shows why they’re an engineer. One of them got me to design a potential divider circuit that they needed (600V in, 3.3V out), and I thought it would be simple enough, seeing as I’ve done it in school before. It turns out that I couldn’t just make up the resistance values I needed: I had to look up E24 resistor values and scale them up and down to fit the ratio I wanted, but then even after that they pointed out that I was using way too much power in my circuit, so I had to account for that as well.
Great first week, let’s hope I don’t burn myself out and get bored of programming and electronics, because it’s really looking like something I’d want to take to uni level!
Fred.
Second week was very number crunchy, and lots of triangles. I finally managed to figure out that algorithm I was working on and I’ve designed an excel document that does all of the hard work for you – you shove your sensor trigger times in, and out comes a bearing to within 5° of the true value. Because of problems I ran into with bearings at specific values (e.g near 0°), it took the entire week to do the algorithm and produce the excel document so that it was accurate. However as much as I loved plugging endless numbers into an excel document and creating numerous cells to process each little bit of data, I’m glad I’m past that stage in my project now. The next step is to code it into Arduino speak, which requires a bit more prowess than I currently have, so the next step is actually to do more research. I’m really happy with how the work is going so far, and I’ve actually managed to get more done than I expected, especially since a lot of it is independent.
As well as electronics and software engineering, Jon has pushed a few buttons and has given me an opportunity to have practical experience in using control theory, where I’ll be writing some code to program a PID loop, specifically to meet the needs of a graduate project that an engineer called Andrew (very good at control theory) is running. Actually I forgot to mention last blog, but when I arrived here on my first day, a few engineering students in their third year of university arrived at the same time, and they’ll be spending a year here before going back to uni to finish their degrees. I’m slightly jealous of their security passes, because mine only gets me around on the first floor, and whenever I need to speak to human resources, I have to get Jon to escort me… Anyway the PID loop isn’t a huge pressure, because apparently it’s only a side project, so it’s not as important as my main one. Also, Andrew said that what I’m doing for him is really only checking his work, as he’s already done most of it and is just giving me a taste of what control engineering is like. It seems like something I would like, especially since it involves a lot of maths, which is always fun.
Going to see if I can find any books on control, or maybe just the maths required. Seems like it could get really interesting.
Fred.
Turns out you can’t just take all the functions you write in excel and just copy them into code, which is really sad. This week it’s been a mix of my control theory side project and coding all of the main project algorithm into the IDE for Arduino. My time here has flown by; I can’t believe that next week is my last week. I’ve been speaking to a guy called Harley who’s here for a year as part of that sandwich degree type thing, and he’s going to be seeing through the graduate project to its finish. Today I sat in a presentation on it, and I got to see all the stages of it’s development, from customer needs to testing it, and there is a lot more control engineering than I first thought. Not only is speed using PID loops, but there are a load of really precise movements that the UUV has to make that require them too. The more I look into it, the more I’m beginning to think that control engineering will be the sort of career path for me. It’s quite exciting to see a few engineers here like Andrew that have only come out of university recently, and are already working on advanced projects and are mentoring placement students.
I really like the look of engineering now, because I was nervous before this placement that it would turn out to be quite boring, but it’s been very challenging and rewarding. I’ve hit a few walls in my project, but it’s really satisfying getting past them, and even more so when I’ve done it without needing to ask for advice. What makes it rewarding is that this explosion-sensing project is actually getting to code something new, so it’s not like I’m able to check my workings against a better copy. If there’s a problem, then I have to either fix it or work around it – I can’t just copy down the write answer and move on. I now know that I definitely want to be an engineer, which is one of the main reasons why I applied for this scholarship, so now I’m focusing on picking up skills and completing my projects!
Projects were finished right on time; I’m so pleased. The main project was a success as I managed to get it working to within 5°, and the PID loop I was working on gave good results when testing it with a potentiometer (acting as the desired speed). I used the graph plotter function on the IDE and got a nice looking graph, which I could change the look of if I tinkered with the constants I set in the PID loop. I’m planning on getting some different examples of graphs and putting them in my scholarship presentation. There was an extension to the main sensor project, which was to make sure it worked if there were sensors missing or broken. I managed to get it working to within 10° if only one of the sensors was failing, but I had no luck in getting it to work if more than one failed… In the end I think my original algorithm depended on using too many of the sensors to be able to deal with losing many of them, but I wasn’t entirely sure, as by then I was reaching the end of my placement and was rounding everything up.
Super happy with the side project, as it wasn’t really going anywhere for a while (because the main project was the priority), but then on the last day me and Andrew sat down and worked on fixing a big instability bug we had and finally broke through! So now he’s working on integrating the loop into the graduate project, making it a really worthwhile project for me.
Seriously impressed with how much stuff I got to work on and do in just the 4 weeks I was here and the scholarship has 100% been worth it. I have lots to talk about in interviews and many new and exciting things to look into – I have been inspired!
You must be logged in to reply to this topic.