Greetings
Hello, and welcome to Junaed Sattar's website at the Centre for Intelligent Machines. I am an alumni of the McGill Mobile Robotics Lab run by Professor Gregory Dudek in Montréal, Canada. I finished my doctoral degree in the Fall of 2011, and have started my tenure as a FQRNT Post-Doctoral Fellow at the University of British Columbia. I am working under Professor James Little, at the Laboratory for Computational Intelligence. This site will remain online for a while, showcasing my PhD research, as I setup my new website at UBC.
My doctoral research was about using visual sensing for human robot interaction, which also relied on tools from machine learning, statistical modeling and human-robot dialogs. This research has been extensively validated on the Aqua family of amphibious robots, in terrestrial and aquatic environments. This page will have information and links regarding my work at the Mobile Robotics Lab. Please use the navigation bar on the left to browse the site.
Latest
Happy new year from Vancouver. My last update on this website, as I slowly settle into life and work in Vancouver. I am now at UBC, working as a post-doc funded by the Fonds de Recherche Quebec: Nature et Technologies; otherwise known as FRQNT (formerly FQRNT and FCAR). Further updates will be on my UBC site, as soon as I get around to set it up.
I have successfully defended my PhD thesis on the 10th of November, and thus have completed all requirements of my doctoral degreee. I guess you can now call me "Doc"!
A couple of new videos demonstrating the visual programming for Ramius (Aqua2) in terrestrial operations (i.e., walk mode) have been added to the videos page.
I have now officially submitted my doctoral dissertation.
As thesis submission nears, I have little time to think about anything else, but here's something maybe worth sharing. Earlier this summer, the French Canadian educational channel Canal Savoir visited us at the lab and recorded a piece for their show called Campus, titled "AQUA -- le robot sous-marin" ("Aqua, an underwater robot" in English). Here's a link to the segment, from their website (in French, of course!).
Just returned from a very engaging and exciting RSS in LA. Had a talk in the Marine Robotics workshop, and it was fascinating to see the other researchers in the domain.
Back from Shanghai, and ICRA. The trip was great, and looking forward now to the end of my thesis work soon. The flight back was long, and the most interesting part was the weird noise coming from the flaps in the wings as the plane took off. Got the cabin crew a bit nervous, and, not surprisingly, myself too!
In Shanghai, China, for ICRA. I am done with my talk, so enjoying the conference and the wonderful city.
Thesis-writing has been taking up all of my time these days, hence the lack of updates. But that said, here is a link to a few video clips that are a direct result of mostly my research -- a couple of these videos demonstrate research from a number of other members of the Aqua team.
Wishing everyone on the occasion of what is now recognized as the International Mother Language Day.
A short video clip of this year's HRI trial in progress can be seen here. This was recorded by the robot's rear-looking camera as the divers were progamming the robot.
.I am the TA for COMP765, Mobile Robotics and Spatial Representation. My own 765 page for this year can be found here.
.Back from Barbados. Eventful trip to say the least, but a lot of fun. Some things worked, some things did not, as is always the case. By far the largest robot presence ever during a trial in Barbados, with three underwater, one surface, and two aerial vehicles being deployed. Now it's time for the post-Barbados debrief and post-mortem.
.In Barbados. Aqua field trip number six. HRI experiment on the agenda, along with a multi-robot rendezvous experiment with an autonomous surface vehicle and Ramius. That's just my own stuff. Lots more exciting things on the cards this year.
.Back in Montreal. Happy to report my sixth ICRA paper in a row has been accepted for publication!
Happy 2011 everyone. I am currently in Dhaka, Bangladesh, and was at New Delhi and Agra in India for ISER 2010. Great conference, great trip. Plus my talk is on tape for the first time ever!
I have officially posted my resume on the site. The LaTeX template was borrowed from Ted Pavlic, who was kind enough to make it freely available.
Sometimes we get questions about our robots, about what they can do and all. Videos of our robots in action are always a nice way to demonstrate these capabilities and our research in the lab. Here's how you can take a look. Go to YouTube, and search for MobileRoboticsMcGill. That's the labs official YouTube username for uploading videos. There are a few there already. Enjoy.
Last couple of months were eventful, and busy as a result. The Aqua team got invited to the G20/G8 summit media event in Toronto, and yours truly, together with fellow PhD student Anqi Xu and Professor Greg Dudek from the MRL has the opportunity to represent the university, and in a broader sense, cutting-edge Canadian technology to the world media during the summits. That generated tons of media interest, evidence of which can be found by a simple Googling of the term "Aqua2". In July, we had the indoor pool trial with Ramius (aka Aqua2) and the new HRI engine, and I'm quite happy to see how the whole system has come together, all the way from the theory to the real-world implementation. We have videos and raw data that I am now going over. There's one more trial before the end of summer, and that's next Monday in our "very own" Lake Ouareau, north of Montreal.
My first web-news entry for the year, and there's a lot to put down. This is a rather personal entry. The last few months have been undoubtedly the most difficult time in my life. I lost my father on the 3rd of February. Mere words won't suffice to describe what he meant to me, so I won't even try. I missed almost all of the Winter 2010 semester, and only now trying to get back on my feet, so to speak. Work was on a hiatus, although I have had a chance to go to Anchorage for ICRA. I'm starting up research again, focusing on getting to the end of my PhD, particularly on a theoretical aspect of dialogue models in a Human-Robot Interaction context. Barbados 2010 was a mixed trip -- some things worked, some didn't. Nevertheless, Considering the dimensions of the experiments I had to run, I am quite satisfied with the outcome. Next few weeks will be invested in theoretical cranking, followed by human-interface and robot pool trials.
Barbados 2010 in just about two weeks time. HRI work in progress.
Have been putting some work in implementing a Cairo-based interface for my Vision-Interaction framework application, something I call VisionSandbox. Yes, it is a brain-dead name, but that's all I could come up with 4 years ago, and it stuck. Anyway, it's written in GTK+, and hence the Cairo enhancements are going in. The goal is to be able to use an overlay as a Heads-Up Display (HUD). The HUD will project information in real-time from the robot IMU, vision systems, depth measurements and so on. All that needs to be written up. On another note, I've been playing around with a 4d Systems micro-OLED display for the robot, and after finally managing to get it to work, have written a C++ class for it for integrating with VisionSandbox. That'll go up on the Code Snippets page soon.
A how-to on creating a custom QNX 6.3 image for installation on CompactFlash cards can be found here.
Haven't had any updates for a while, but lots of things have been going on. Have started to look at a quantitative model for HRI with uncertainty in the fray. Among hacky things, wrote a driver for Nomad Scout robots in Player to work with our sole-surviving Scout, and it looks okay so far (largely based on the code for the Nomad 200). I intend to post it in the code snippets page, after some debugging and cleaning up. Also looked at a toolbox called Gearbox, which is required to run Hokuyo laser range finders (in Linux, in my case). There was a checksum bug that prevented our particular model(UTM-X001S) from working, I got around that with a really ugly hack. As soon as de-uglyfy that code, I'll post it here too.
I thought this came out pretty cool!
Somehow, I ended up with the best robotics paper award in the Canadian Conference on Computer and Robot Vision held in kelowna, BC! That was a pleasant surprise to say the least. The said paper will be added to the MRL publications page soon, if anyone's interested.