Autonomous systems postgraduate personal statement example

Personal statement example

My interest in autonomous systems began less with self-driving cars than with a corridor in my university's engineering building. For my final-year BEng project I built a small differential-drive robot with a low-cost lidar and wheel encoders, and set it the task of producing a usable map of that corridor. The mapping worked reasonably well in the middle of the floor and fell apart near a glass-fronted display cabinet, where the lidar returns became unreliable and the pose estimate drifted badly. Tuning the filter helped a little; understanding why the sensor model was wrong helped much more. That distinction, between adjusting parameters and reasoning about what a system believes and why, is what I want to study properly at master's level.

My degree gave me solid foundations in signals, control and embedded programming, and I chose electives in digital control and machine learning. I wrote my controller code in C++ on a microcontroller and used Python and ROS on a laptop for the mapping, which taught me how much of robotics is unglamorous integration: timing, message rates, and coordinate frames that quietly disagree. I read Thrun, Burgard and Fox's Probabilistic Robotics alongside the project and found the treatment of probabilistic state estimation clarified what I had been doing by trial and error. I am now keen to go further into planning under uncertainty and into the verification side, since a system that behaves well in a lab corridor is not the same as one that can be trusted in a public space.

Alongside my studies I have worked part-time for three years as a mechanic in a bicycle shop. It is not robotics, but it has been useful preparation in ways I did not expect. Diagnosing a fault on a customer's bike means forming a hypothesis, testing the cheapest one first, and resisting the urge to replace a component because it is the part you understand best. I also have to explain to people why a repair matters, in plain terms, without being condescending. I think engineers working on autonomous systems will increasingly need that skill, because public acceptance of these systems depends on honest explanation of their limits.

At home I share responsibility for my younger brother, collecting him from school several afternoons a week and helping with homework. This meant my final year ran on a fairly rigid timetable: lab time booked in advance, evenings used for writing rather than debugging. I learned to break the project into pieces that could be finished in ninety minutes and to keep a written log so I could pick up a problem where I left it. I mention this because it explains how I work rather than as a difficulty; I am organised largely because I have had to be, and I expect a demanding taught master's to require the same discipline.

I am particularly interested in the gap between simulation and reality. In simulation my mapping worked almost immediately; on the real floor it did not. I would like to study how uncertainty is modelled and propagated, how planners handle incomplete information, and how autonomous behaviour can be tested in a way that gives reasonable confidence rather than a single successful demonstration. In the longer term I would like to work in the robotics industry on inspection or logistics systems, where the environment is structured enough to be tractable but varied enough to be interesting. A taught master's with a substantial project is the right next step: I want a second attempt at that corridor, with better tools and a clearer understanding of what I am actually asking the machine to believe.

Review this personal statement

Latest reviews