Over the last six months I’ve been fairly busy with work, but also with a number of personal projects. I’ve been to Spain and California, and have continued to cycle the roads of Wiltshire. However, there hasn’t been much that I considered worthy of writing about here.
This entry will summarise what I’ve been up to and link to some of my project work.
When the weather was warmer I cycled quite a lot. My fitness improved to the point where I was as fit as I was at my peak last year, before falling ill. So far this year I have ridden 1860 km, with July being my best month at 543 km. And with an overall average speed of 27.2 km/h I have been quite pleased with myself.
The approach of winter has made motivation a problem for continued cycling, but I have noticed a correlation between the amount of hard exercise and my general well-being. As that in itself has been a big challenge since I was in hospital, I should try to make an effort to keep physically active throughout the winter.
In August I travelled to the north of Spain with my parents. We stayed in the city of Pamplona, famous for its bull run, and ventured out into the Pyrenees mountains for day trips. Despite our car suffering a double puncture at the top of a mountain somewhere on the French-Spanish border, it was a very nice week.
Between September and October I spent three weeks in the USA with work. We were based in the Mojave desert but managed to spend our weekends away. We went to San Francisco for three days and visited Las Vegas for two days. For me, San Francisco was the definite highlight and I am keen to go back there at some point in the future.
Several projects have occupied my free time recently, and probably too many to progress if I’m honest.
In the past few years I have been fascinated by the idea of an autonomous aircraft, a UAV or drone as commonly known. Although they can now be purchased and used almost straight away, my interest lay deeper. I always want to know how something works and why it was designed that way, and I thought I could embark on a journey to learn exactly that. Aircraft are expensive to learn with, however, as crashes are likely and costly, not to mention dangerous. I therefore decided I should start with a ground-based robot.
I quickly purchased some hardware and started programming an Arduino to control my robot. It was refreshing to include hardware into my projects as a lot of my previous work had been exclusively software.
Once I get a robot built that can successfully navigate and avoid obstacles, there are several potential avenues to explore. Ultimately my goal is to have a number of vehicles that can manoeuvre autonomously but that have the ability to build a mesh network and transfer data between themselves and some sort of base station. The application I had in mind was to explore a disaster area — many low-cost vehicles could be sent to explore and maybe map a dangerous area. And they could carry a payload of sensors to determine different sorts of hazards. It might also be possible for them to construct a temporary communications network to augment damaged or overwhelmed infrastructure.
I have started to document my ideas and Arduino source code on GitHub, accessible here.
As the robot project developed, the idea for flight didn’t go away. I bought a cheap micro-quadcopter and had a great time flying it around the office. This inspired me to build a bigger one, and that’s what I did in October. It was built using off-the-shelf components and I’m still trying to tune the control system, but it is very fun to fly provided I have enough space.
Rather than just build and fly a quadcopter, I’ve also been trying to understand how it is controlled, and refresh my poor knowledge of control theory in the process. And I also have an aspiration to link a future autonomous quadcopter to my ground-based robots to further increase the possibilities of disaster area surveying and sampling.
I have been completely won over by the IPython Notebook concept recently. All of my learning and development now takes place within a notebook, and I store most of them on GitHub. Together with Python libraries such as NumPy, SciPy and SymPy, the support of LaTeX and the ability to view notebooks in the browser using nbviewer, I can document my work with a mixture of formatted text and equations, Python example code and charts of results.comments powered by Disqus