Hey everyone! So it’s the first week of September and I’ve just completed my first week of my second year at USC and the week before that I wrapped up my robotics internship at Facebook where I worked on a research and development team! It’s been a hectic few weeks, from de-orienting from Facebook and moving out of my summer apartment to moving into my new apartment at West 27th Place and fiddling around with my class schedule and deciding the various commitments that I want to make this upcoming year!
After getting back to school, one of the first things that I did was return to the Interaction Lab where I had worked during the Spring Semester to catch up with everything that’s been happening over the summer! The project that I am on is a proxemics engine for Human-Robot Interaction. This software backend gives users the ability to ‘automate’ the positioning of the robot during an interaction, whether it be telepresence or kiosk mode, allowing for a large variety of interesting use cases such as multi-presence and full-duplex telepresence communications. As this project is wrapping up, I need to consider my future in the Interaction Lab and what I want to study.
One of the most impactful things that I’ve learned since getting to USC has been the importance of always keeping an open mind and giving everything that you do an opportunity to relate to each other and impact each other in a positive way. At Facebook, I was given the opportunity to meet and work with a lot of different people with a variety of technical backgrounds and one of the coolest projects that I had an impact on was related to the application of cutting-edge machine learning techniques to social webs and the underlying concept of that project was that each individual user is a different entity.
By tying my previous work in proxemics modeling in Human-Robot Interactions as well as my newfound interest in adaptive machine learning models, I think that I would like to explore my own research project that develops adaptive models for holistic interaction informatics!
What this basically means is that, if successful, I will have created a model that controls various important aspects of a human’s interaction with a robot—volume, physical location and orientation, light displays and screen displays, even emotional state of the robot to optimize the quality of the interaction and the ‘goals’ of the robot. Essentially, I will be able to give a metric of ‘quality’ of model as a function of the ‘amount of interaction’ (defined by the integral of the quality of the interaction over time). This has a very large range of potential impact not only in the research world of robotics but more importantly in the practical applications of robotics in society and in the home, giving users and developers an easier way to control the quality of interaction and providing an avenue for seamless integration of robotics into the general population.
So I hope to get on this project ASAP and hopefully have something to show very, very soon! Also, with Football season around the corner–I’m very excited for kickoff on Saturday! Fight on!Meet Eric