Time to apply for PhD programs again. Maybe I’ll have more luck this year. For those interested, here is my statement of purpose. Let me know what you think.
I am looking at studying in the field of Computer/Human Interactions, but with a special refinement. I want to look specifically at robotic systems and possibly virtual reality environments and how they interact and react to the user. It’s no question that these fields are growing and that they will soon be commonplace, but I have yet to see many diligent studies at this interaction. Most require the use of external graphical user interfaces with a few employing only rudimentary gestures. In virtual worlds, this can be even more difficult since employing a user interface in any sense can distract or destroy the immersion of the simulation. To combat this I would like to look at technologies as they exist today as well as the latest emerging technology, with CMU’s own robotics group being on the forefront of this field, and focus on current sensors and intentions and how they can be merged into a common control structure.
This interest of mine has come directly from past experience on how these limitations can affect and change lives. I have served in the U.S. Air Force for six years with a good portion of that time spent overseas in Iraq. My role was to operate the defensive vehicles for convoys that travel through hostile zones. During this time I have seen some of our advances in robotic technology and it has generally been positively received. However, I could imagine having the ability to ‘summon’ our vehicles via a common control structure and have them shield or warn of danger. Self-driving vehicles are coming, and due to the nature of moving supplies and equipment in conflict zones, I’m sure they will be adopted quickly. However, they will still require protection and thus there will still be boots on the ground with the vehicles. If some interface existed between these systems and the troops, it could save lives.
In the limited personal time I have, I have attempted to build and define exactly what would be required for such a system. I have been working on projects that involve using multiple sensors to track and understand the user in a way that can be quickly and easily interpreted but is also natural to the user. I have one such project with multiple microphones in an array around the robot and a camera watching the user to elicit input. In my search, I’ve been surprised at the lack of resources available to the community in general. Some common approaches, such as audio beamforming, or using multiple microphones to isolate and capture sound are woefully inadequate or incomplete in the public domain. Others, such as the available computer vision libraries are still maturing and could be advanced specifically for this application.
The research, as mentioned before, would also apply well to the virtual domain. Robotics use sensors to detect what is happening in the external world and attempt to understand and react to the user’s wishes. Likewise, virtual systems are looking at the user for the same reasons. The difference is that it may be possible to provide some hints using the visual/tactile systems of such devices. Such feedback may be passed on via augmented reality systems to robotic systems in the future if such methods are deemed superiors to non-augmented approaches.
I also have past professional experience in some of these fields such as voice recognition and device automation. I have worked for six years following the military at Tellme, a company that focuses on voice recognition and automation. This company was bought by Microsoft and its technology is what powers the Cortana personal assistance as well as projects like the Ford Sync and Xbox Kinect Voice Assistance. I know the difficulties and effort required for developing such systems and believe it would be an asset to my research.
Which brings me back to why I am seeking a Ph.D. For a long time this has been a hobby and passion of mine, but my time and resources has always been limited by my external forces. I have not had the formal training or the credentials required to work in a field that would coincide with my research. It is my goal to gain both of these so I can work on a solution for not only current systems, but also systems that are not yet even imagined in the future. It is my goal to be a driving force in this field.
The school of choice for me to accomplish these goals is Carnegie Mellon University. Not only do you have an amazing robotics program which I could communicate with and draw inspiration from, but you also have many accomplished and inspiring professors in the area. Professor Pei Zhang had done some amazing work in the area of Mobile Robotics and I would love to work with him in new controls systems and feedback mechanisms. I would also like to work with Professor Ian Lane in his field of Human Computer Interaction as well as Professor Abe Ishihara in the field of Control. I feel my research would touch on all of these, and any would be an ideal mentor for my studies.
Finally, while my goals and passions have been laid out plainly in this letter, my real goal is to learn how to conduct this research and how to document and report it for the use of the general public. Even if I work on similar projects, working for my Ph.D. will give me the skills and credentials required to work in this field in the future. I am first and foremost looking to be a student and do what is required to become a master in this domain.