Entries : Category [ Education ]
[Miscellaneous]  [Computers and Technology]  [Travel]  [Education]  [Hacks]  [Robotics]  [Science]  [Programming and Software]  [iPhone]  [Digital TV and Video]  [Intellectual Property & Copyright]  [Personal] 

01 April
2007

I am teaching a computer software development course. So far the students are working on a web server development project with some very advanced features. While on a research trip to the US last week, I managed to swing a deal with a major ISP to purchase all of the working assignments from those students who are willing, for a whopping $440,000 USD plus royalties. The exact revenue to be returned per person is hard to predict since our university will probably keep about 95% according to their IP policy, but that should still leave an interesting revenue per person.

The students need to vote on the proposal, get a 85% majority, and provide their source code. Then a team of team get to accompany me to Orlando at their own expense to provide the needed deployment team get get everything running and integrated into the ISP's web site. The amazing part is that the deployment team gets a bonus of 200 pounds of coffee beans and deodorant to allow them to get the job done on a very accelerated schedule.

The whole package was finalized today, April 1st.


By Gregory Dudek at | Leave a comment |    
02 April
2007

Just back from Orlando where I gave a colloquium talk at the University of
Central Florida (I talked about the AQUA robot and some of the vision
algorithms we are developing for it). [This was a second trip after a botched visit recently.] There is a very active computer
vision group there that is really
agressive and productive with many interesting projects on the go. Among other things
they just got a nice DARPA contract for a fleet of flying air vehicles (UGV's) to complement the one they already have.

I took three hours off and made a quick trip to the Kennedy Space Center on the coast.
They have several old rockets and shuttle mock-ups on display. It's kind of interesting, but I am not sure it's worth 2 hours driving and the $35 admission price if you onyl are left with a
single hour to visit. On the other hand, there are various included movies (including IMAX) that I wasn't able to see, and they seem to be key parts of the site. A much more reasonable plan is to take the bus tour(s) of the entire site, but that is both time consuming and requires advance reservations. I did that years ago and found it very worthwhile, plus there is interesting wildlife to see occasionally (such as alligators).


By Gregory Dudek at | Leave a comment |    
30 April
2007

Last Friday I was at the University of Texas at Austin for the doctoral defense of Mohan Sridharan. Mohan's thesis dealt with color recognition and constancy (and re-calibration) for navigation and localization.

A key part of his work deals with adapting to illumination changes. This is accomplished by using a hybrid representation of both color and illumination change, and using different mechanisms to deal with gradual versus sudden changes. He passed comfortably.

The robotics labs there were large and very active. I finally got a chance to ride on a Segway myself, which was surprisingly easy and smooth. Most Segways one finds in robotics labs are not suitable for human riders.



By Gregory Dudek at | Leave a comment |    
14 November
2007

McGill University managed to break into the top dozen schools in the world, according to the Times Higher Education Supplement (THES). This is one of the more rigorous and large-scale university ranking exercises, which spans universities from all over the world.

The result is a ranking of universities the world over based on research related criteria. This is rather different than ranking system than used on the Canadian MacLeans survey and several similar surveys where most of the emphasis on highly subjective or hotly debated factors that are thought to be important for undergraduate education. The objective of the rating was that 60% of the overall rank is is awarded based on objective quality indices for research faculty. International reputation based on a survey of other academics accounted for 40% of the score, and 20% was based on citations as measured by Thomson ISI's database. Of course, the specific people used for the indices are a source of bias, and one criticism is that this survey includes business leaders, which some academics find objectionable.

No other Canadian school ranked as highly, but of Canadian schools the University of Toronto and UBC both did well (45th and 33rd places). Schools that did better than McGill include Harvard, Cambridge, Oxford, Yale, MIT and CalTech. Some pretty distinguished places did worse. The rankings cover all areas, so schools with a more specialized focus had a handicap. For example, Tokyo University came in 17th, but I believe it did better in the sciences that in the overall ranking. Carnegie Mellon University came in 20th, but is almost always ranked in the top 3 for computer science.


Of course like all such rankings there is plenty of debate over the precise criteria used and their relative weights. No doubt the precise ordering would vary if the relative importance of the criteria were changed, and many alternative re-weightings are reasonable. For example, National University of Singapore well from 19 to 33 this year, but I doubt the school itself changed that fundamentally. UC Berkeley also fell quite a bit (to 22nd place). Whatever flaws there are, this is a pretty solid ranking system, preferred methodologically to the ranking in, say, US News and World Report (although I have to confess that USNews results for computer science are quite reasonable, with Carnegie Mellon University, Massachusetts Institute of Technology, Stanford University and University of California–Berkeley being tied for top spot).

In summary, the numbers are interesting and it's nice to see McGill moving progressively upwards in successive years (in 2005 McGill was ranked 24th, which was still pretty impressive).


By Gregory Dudek at | Leave a comment |    

These charts show recent job trends, and computer science is trending up. So is C, as well as Python. There is also a chart showing what programming languages are mentioned explicitly. Surprisingly, C remains a dominant theme. Python has been exhibiting a clear upward trend, but it's hard to make out on this figure since the absolute value is pretty small comared to C. This data is based on searches of job postings, so there is room for debate for its depth and reliability. For example, I suspect that many "high end" jobs don't bother to mention specific programming languages, yet there are doubtless expectations regarding what language will be used in a workplace. For example, we use Matlab in my lab, but I never bother to mention it as a selection criterion.

job postings with various programming language specifications

The following chart shows the frequency of occurrence of various academic background phrases. Note that the absolute number of jobs may not be the same as the ease of getting a job, since there are both demand and supply factors in the marketplace. If there is only one person with a background in statistical anthropology, just two job postings might be more than enough.
Another factor to take into account is bias in the sources used to compile this data. I don't know how they stack up, but the figures should be taken with a grain of salt.


"computer science",  "electrical engineering",  "software engineering", physics, math Job Trends graph



By Gregory Dudek at | Leave a comment |    
18 March
2008

The St. George's School Robotics Team, with Nicholas playing a key position, recently participated in a regional robotics competition. They built a robot that had to drive around under remote control and shoot bean bags at a large (30 foot high) score board. Their robot had an uncannily good throwing arm. In addition, they also designed and produced an shiny exhibition booth that featured a working animatronic flying dragon, and descriptive video (in two languages) and associated materials.

Unfortunately, they experienced an electrical short early on in the competition which badly damaged their hardware, and the robot was not able to perform well. That's how so many technology demos go. Nevertheless, the whole thing was very impressive. A big high profile competition like this is a great way to stimulate enthusiasm for, and interest in, science and technology.

The robot itself operated using a differential drive steering mechanism: two independently powered real wheels. Two small front wheels provided balance. As is often the case with differential drive, good straight-forward steering was tricky, but they addresses this by using a pretty large wheel diameter and comparatively low torque.


By Gregory Dudek at | Read (1) or Leave a comment |    
Prev  1   [2]   3   4   5   6   Next