Tag Archives: tcs

INFORMS on the Smart Grid

I hadn’t heard of the “smart grid” until I arrived in Oregon.  Our department is pushing for a sustainability research collaboration initiative, SENERGI, and so it wasn’t long before I heard our former director, Terri Fiez, talking about the smart grid.  Now, at INFORMS in San Diego, I’m listening to a keynote on the smart grid by Richard O’Neill, the Chief Economic Advisor to the Federal Energy Regulatory Commission

For the unenlightened, the smart grid is the idea of changing the price structure of electricity as well as the appliances that use electricity to manage congestion.  As we move toward more electricity use (i.e. from gas-powered cars, to electricity-powered cars) and electricity generated from renewable and time-constrained resources, congestion could cause more frequent brown-outs than we’ve seen.  Some concrete examples of “smart” include:

  • Appliances equipped to decide when it is best to run: refrigerators that turn off and on to minimize draw on the grid during high-demand hours, dishwashers that wait for low demand hours to run.
  • Batteries equipped to charge by use time: if you arrive home from work in your car (which I’ve been told is possible, but have yet to experience), and plug your electric powered car in along with everyone else who works 9-5, but you don’t need your car again until 6 AM, then the battery will decide to not charge until the middle of the night, perhaps according to some neighbourhood schedule.
  • Batteries used as storage devices on the grid: if you don’t use your car during the week because you do walk or bike or bus to work (congratulations), then the grid could use the battery as a storage device on the grid, charging during low-demand hours and discharging during high-demand.  (Of course, if you don’t need a car during the week, consider not owning a car.  Renting a car almost every weekend is often cheaper than owning a car.)

Of course, for such a system to work, there are significant engineering (after all, even my dishwasher’s simple “delay-start” timer doesn’t work), design and optimization challenges.  The market will become much more complex – will every appliance be considered a player on the market?  Sheesh!  Currently for the (much simpler, I imagine) pricing problems, the cost functions are linearized, which apparently isn’t a great approximation – essentially treating AC current as DC current.  In a system that is worth $10^12/year, a 1% savings is huge news.

I worry though … what if my laptop tells me I can’t write an email at 2AM because I woke up in the middle of the night wracked with algorithmic thoughts because my battery has been discharged so my neighbour can run their washing machine.

How to find a postdoc

While I hardly think I should be doling out advice …

In algorithms, there have been a lot of postdoc positions advertising on the two main email lists, TheoryNT and dmanet.  In my experience, many of the positions are in Europe.  I’ve found that a lot of postdoc’s get their position by word of mouth.

I think, by far, the best thing is to get a postdoctoral fellowship.  Freedom!  It seems NSF doesn’t have a fellowship program for people in computer science.  (Is that actually true?) But I have seen (and ignored, as I am not an American citizen) plenty of postdoc fellowship programs for Americans.  If you aren’t American, try your home country.  NSERC has great fellowships for Canadians that you can take out of the country if you got your Ph.D. in Canada and is tax-free if you take it to McGill.  The short of it is, if you have a fellowship you have the academic freedom to study what you want to study.  You can work with whoever you want, whether or not they have a research grant to pay a postdoc.

I’ve also thought that if you plan far enough in advance you could contact someone you really want to work with and convince them to write a grant with your help that includes funding for a postdoc.  Any thoughts of whether that would work?  I know NSF now asks for an “advising plan” when requesting funds for a postdoc salary.  Would having the potential postdoc involved in the writing process help?

And there are schools and departments that have their own postdoc program – I think U. Penn and U. Toronto do.

Any other suggestions?

What theory should every non-theory Ph.D. student know?

I’ve survived my first week of teaching graduate algorithms and data structures. “Survived” really isn’t the right word. I’ve had a lot of fun and the students in the class are bright and interactive, which makes a 50 minute lecture go by in a flash.

Since the time is going by so quickly, I realize the need to consider more systematically what should be taught in this course. As you might know, I am the only algorithms/TCS person in the department (who isn’t emeritus) and so I will likely be able to quite easily affect the graduate curriculum.  I’m impressed by the amount of theory that the CS Ph.D. students are required to take here (at least in comparison to Brown, a theory-heavy school).  Each student must take the (10-week long) courses called “Algorithms and Data Structures” and “Theory of Computation and Formal Languages”.  Beyond that, there are several other optional algorithms and complexity courses offered every second year.

There has been some discussion on My Biased Coin on what every theory Ph.D. student should know. My question is: given 20 weeks of class time (three 50 minute lectures a week), what topics in TCS should every CS Ph.D. student know?

Why blog?

Bill Gasarch asked me to make a statement about my blog and in responding to him, I realized I might as well post it here.

This blog will likely be YATB (yet another theory blog) – hopefully I will have something new and interesting to say.  I’d been subscribing to the Theory of Computing Blog Aggregator for some time now (a tool I am very thankful for), but of the 20 or so blogs it contains, Sorelle Friedler’s is the only one by a woman.  If we want to balance the gender inequity in our field, we need to get more women into the system.  If we raise the profile of the women already in TCS (or CS or math or engineering), then perhaps it will seem more desirable to undecided female high school and undergraduate students.

I can’t remember the reference, but someone pointed out on a TCS blog that a particular program committee was rife with TCS bloggers, so yes, my motivation is also selfish.  I am hoping that this blog will gain me exposure, particularly in TCS, and garner me advice.  After all, I’m on my own out here.  Already it feels less lonely.

So I’m not sure what I will blog about.  It will likely be a mix of technical posts and posts about my professional life.  I will feed the TCS-related posts to the ToC blog aggregator.

Algorithms advertising for incoming grad students

One of my first professorial tasks will be a 4-minute talk to incoming grad students. I don’t expect any of the students in the audience will be explicitly interested in algorithms research.  After all, when they applied, there weren’t any algorithms profs here to advise them.  So, I’m not sure what the best use of this time is.

I could ignore the fact that they are likely interested in doing everything but proofs and give my 4 minute blurb as though there were potentially interested students in the audience.  I could appeal to the coadvising role of “whatever your research is, there is probably an algorithmic viewpoint”.  I could just say “this is what algorithms is” and leave it at that.  Besides, (and lucky for me) all the CS Ph.D. students have to take a grad algorithms class – and guess who is teaching that in the first quarter.

However, as far as I know, the EE and masters students don’t have to take the algorithms class.  Do you think I should be convincing them to?

Newly minted

I just started – a week before my official start date – as an assistant professor at Oregon State University in the School of Electrical Engineering and Computer Science.  I feel very luck to be here, to have landed a job in a year when many of few job openings were rescinded because of the crumbling economy, let alone in an incredibly friendly department in a beautiful town near mountains and oceans.

I am alone as a TCS person in the department, and I’m sure I will feel some isolation as a result.  I am really hoping that I will be able to work on some interesting problems with non-TCS people.  My background, though, is in very traditional algorithmic work, so I am sure there will be growing pains. I’ll start with what will probably be the first of many requests for advice: any advice?