16 November
2008

We got our company Independent Robotics incorporated, and had it's name officially registered.


Posted by dudek at November 16 17:59 | Leave a comment | permalink link to this entry |
27 October
2008

Apple Time Machine is an attractive backup system for OS X, but I had a hard time fixing it after it recently stopped working for me after I accidentally unplugged my firewire backup disk during a backup operation. For subsequent backup attempts, I got the error message that the backup had failed because a "time machine error -- Unable to complete backup" had occurred. Not so helpful.

By running the Console application, I was able to look at the error messages from backupd. In my case, there were some files that could not be inspected leading to messages like this:


... ...
There's more. Read the whole story on "Fixed Errors with Apple's Time Machine Backup System"
Posted by dudek at October 27 23:40 | Leave a comment | permalink link to this entry |
24 October
2008

The Globe and Mail newspaper released their Canadian University Report. Like all such surveys, the results need to be taken with a BIG grain of salt, in part because it was based on students judging their own university (so there is a lot of selection bias). Similarly, note that most Americans who have never been outside the USA think it's be best country to live in: what does that really tell us?

Still, it's nice that McGIll University did well. Also interesting is the fact that Canadian students everywhere are generally satisfied with just 7% saying otherwise. This is consistent with my longstanding belief that the Canadian education systems works really well from both a social and personal point of view.

McGIll is most naturally compared to the University of Toronto, University of Alberta, UBC and Waterloo based on the size, diversity, scope and impact of the schools. This is especially true in Computer Science.

McGill is top in this set of 5 in the important areas of:


academic reputation,
campus atmosphere,
career preparation,
campus technology,
libraries,
student residences,
most satisfied students,
quality of education,
quality of teaching.


... ...
There's more. Read the whole story on "Globe's Canadian University Report: McGill, Toronto, Waterloo, UBC"
Posted by dudek at October 24 13:17 | Read (1) or Leave a comment | permalink link to this entry |
22 October
2008

This post discusses how to distribute music to a collection on locations around a single house, but it works equally well for distributing music across the internet since it involves creating an internet-based streaming "radio station" using icecast, as well as Airtunes.



iTunes being streamed in from icecast




Last weekend my wife Krys threw a party for me. Like any party you have yourself, it was a bit frustrating since there is so much greeting and running around to do that I didn't get much chance to spend much quality time with anybody, and everybody there was somebody I would have liked to sit and talk to.

Party
Party



I had hoped to get people dancing (and failed), but as an aside I arranged to broadcast music around the house to several locations on 2 floors. To do this, I used iTunes (on a Mac) to produce a playlist. The problem then, was how to distribute a synchronized playlist to at least 3 locations within the house. As an aside, I wanted to use icecast (i.e. shoutcast) so I could do other stuff as well like (later on) broadcast outside the house. The Shoutcast protocol can be used to broadcast to Macs, Windows, Linux, FreeBSD and Solaris (and more). Shoutcast is freeware while the compatible Icecast (or Icecast2) is true GPL (and I believe it was released before Shoutcast became free).

Option 1: Use Airport Express: The easiest way to do this is to use airport express wireless "bricks". You can broadcast to these easily from iTunes, and it it just works. My problem was that I didn't have enough Airport Express bricks; some locations had to be actual computers, but iTunes does not support broadcasting to another computer.

Airtunes Setup within iTunes





Option 2: Use Airfoil: There is a software package called Airfoil. This allows you to capture the audio output (from any program) and broadcast it to an Airport Express unit, or to a mac or Windows computer running their free Airfoil Speakers application. It's very nice and only costs $25. Being a bit of a hard-core hacker, I was interested in an icecast solution I could roll myself, for several reasons including the fact it would allow me to subsequently broadcast content over the internet, for example to my office.

Option 3: Roll my own Icecast2 server: this is obviously what I chose to do. It turned out to be a lot harder to figure out than I expected and had I been less reluctant to admit defeat and more practical, I would simply have purchased Airfoil. On the other hand, it's pretty easy once figured out and I needed some hacking-based stress relief. Read on for the recipe.


... ...
There's more. Read the whole story on "Broadcasting music to multiple speakers"
Posted by dudek at October 22 21:12 | Leave a comment | permalink link to this entry |
07 October
2008

One of the disadvantages of underwater (or amphibious) robotics is that there is a lot of logistic overhead to do real experiments. Just doing a smallish test in the pool requires not only that the whole pool be reserved (which is hard to accomplish), but that a fair bit of equipment as be transported and installed poolside.

In practice, this usually involves a team of 6 to 8 people, including swimmers (or divers) who accompany the robot and take photos underwater. Since we rarely get more than a couple of straight hours of pool time, this means we needs to minimize setup and take down times. When the robot is tethered, just managing the delicate fibre-optic cable is a hassle, and spooling it up wastes valuable time. Using a more robust cable would mean a heavier cable that would interfere with the robot's behavior and the measurements of its dynamics.

In the last year or so we have been doing lots more experiments without any tether. This has been a big win, and setup time (and logistic support) is now pretty limited. That's great, but it a small way I miss the bug party atmosphere of a big crew.


Posted by dudek at October 07 18:17 | Leave a comment | permalink link to this entry |
01 October
2008

Last week I was at the International Conference on Robots and Systems (IROS 2008) in Nice, France. This is one the the two huge annual conferences on robotics research. This one places particular emphasis on complete systems and it has an especially large representation from Japan and Asia, being sponsored by the Robotics Society of Japan (RSJ) as well as the IEEE.

As I get older, I end up spending a larger fraction of my conference attendance time having discussions in the hallways and making plans, instead of merely attending the technical sessions. Despite that, I heard a variety of talks spanning both areas I work in like SLAM and underwater robotics, as well as topics like the Design of Humanoid Robots that I only track via conference presentations.

One of our own presentations was an overview of the Aqua robot project, and how we are moving to increasing levels of autonomy with the underwater vehicle. The presentation was tricky since rather than focus on a single narrow technical problem, it had to weave together a group of projects and problems that, together, allow the vehicle to operate semi-autonomously. In the same conference session on underwater robotics we also heard about a few types of autonomous surface vehicle and how they can navigate autonomously.

Dave Meger presented work he had done with Yiannis and myself as part of his Master's thesis. That dealt with using a mobile robot to help the nodes of a sensor network to estimate their own positions. In particular, as the robot explores the metric embedding of the network, and can select various path planning strategies, and these directly impact the accuracy of the localization process.

Acropolis Conference Center in Nice, France
Acropolis Conference Center in Nice, France



Shadow Hand, IROS, Nice, France
Shadow Hand, IROS, Nice, France



Posted by dudek at October 01 14:20 | Leave a comment | permalink link to this entry |

The Large Hadron Collider is scheduled to have it's first actual particle collisions soon, as opposed to just being "warmed up".

[ I wrote this article a while back, and had it cached until collision time. I didn't expect, and the time, that the LHC would get so much coverage. ]

From the LHC web site:

There are many theories as to what will result from these collisions, but what's for sure is that a brave new world of physics will emerge from the new accelerator, as knowledge in particle physics goes on to describe the workings of the Universe.

There are some people who believe that is a particle accelerator that will be able to create small black holes. This would be truly microscopic and not the same as Einstein black holes (they would be much smaller than a neutron).

These black holes should evaporate quickly and be no threat, although there is a minority of people who fear it could destroy the world if the black holes persist. The seems unlikely since cosmic rays have as much energy as the LHC beams, and they haven't destroyed the earth yet. On the other hand, cosmic rays have very high velocities while the LHC could produce occasional low velocity black holes. Thus, this does not appear to be a very credible threat. More importantly, there are other cosmological arguments to suggest such non-evaporating black holes cannot exist. On the other other hand, it should be a good excuse for an end-of-the-world party. There's another end-of-the world mechanism as well, based on strangelets made of special quark matter, but it too seems to be ruled out and I don't have the stamina to explain it.

With respect to data processing, the LHC is supposed to produce 15 petabytes of data per year (i.e. (15 thousand terabytes -- 15,000 trillion). The entire Google index of the web, in comparison, has processed one trillion links as of mid 2008 and index a far smaller number (according to the Google blog).

Live picture. CMS construction at Point 5. Click to enlarge and refresh.

The full story on why the LHC is considered safe can be found
at the CERN web site itself.


Posted by dudek at October 01 00:00 | Leave a comment | permalink link to this entry |
11 September
2008

Every times I see it, I like it more.


In The Know: Are We Giving The Robots That Run Our Society Too Much Power?


Posted by dudek at September 11 00:00 | Leave a comment | permalink link to this entry |
05 September
2008

One of the main things that makes me love the iPhone is the combination of a good solid web browser with the phone functions. I once had a Treo and there is no comparison at all. Sometimes people complain about the browsing speed, so it was interesting to see the side-by-side comparison below between the Apple iPhone and the RIM Blackberry Bold.

The video makes and interesting and compelling demonstration, but the upshot is that the iPhone is a lot faster. I wonder if this is partly because the iPhone is based on a robust UNIX-like kernel which is the backbone of the internet, as well as because the Safari-based browser has been the subject of development for the huge Mac community for many years. (Claims regarding the speed of Safari version 3 in comparison to other browsers have been around for some time.)


Posted by dudek at September 05 00:00 | Leave a comment | permalink link to this entry |
03 September
2008

Today was my second day in my new role of Director of the McGill School of Computer Science (roughly equivalent to "Chairman of the Department"). The school is part of the Faculty of Science, but it has many links to Engineering as well. I also have some plans to increase the visibility of robotics at McGill, and this platform should help.


Posted by dudek at September 03 18:27 | Leave a comment | permalink link to this entry |
31 August
2008

On November 14th, 2008, Michael Geist will give two talks in Montreal. One will be at the Computer Science colloquium at the Cchool of Computer Science of McGill Univeristy, and the abstract is below. The other talk will be at the Trudeau Foundation. The McGill talk is open to the public; I don't know much about the other talk.

Why Copyright? The Fight for Canada's Digital Future"

Abstract:

In June 2008, the Canadian government introduced Bill C-61, new
copyright legislation that closely followed the U.S. Digital
Millennium Copyright Act. The public response to the bill was both
immediate and angry - tens of thousands of Canadians wrote to the
Minister and their local Members of Parliament, leading to town hall
meetings, negative press coverage, and the growing realization that
copyright was fast becoming a mainstream political and policy issue.
The "Canadian copy-fight", which includes many new advocacy groups and
the Fair Copyright for Canada Facebook group that has over 90,000
members, has attracted considerable attention from the mainstream
media, with many wondering how copyright had emerged as a contentious
policy issue. This talk will assess both the legislative proposals and
the Canadian copyfight experience in an effort to answer the oft-asked
question – "why copyright?".


Posted by dudek at August 31 14:41 | Leave a comment | permalink link to this entry |

Earlier entries...



Science Blogs - Blog Top Sites Science Blogs - Blog Catalog Blog Directory Science Blogs - Blog Top Sites Science Blogs - Blog Catalog Blog Directory