Ask HN: Are There any Interesting Jobs in Traffic Engineering for a CS Major?

https://news.ycombinator.com/item?id=8000650

by learc83 • 12 years ago

8 8 12 years ago

news.ycombinator.com

I've got 2 semesters left to finish up my CS degree. I'm 30 and I've got about 8 years experience as a freelance programmer.

I'm interested in solving problems related to traffic engineering. I don't have the time/money to go back to school again to get civil engineering degree, so maybe something like working on traffic simulation software?

hga• 12 years ago

One possibility is companies that make electronic devices for this; not all of it is classical embedded, BTW, or at least so I remember from one small company I interviewed at which used Visual Basic in the '90s and at least a bit beyond, although I can't remember exactly what they were doing. That won't be traffic engineering per se, but will be a part of it.

Various units of government with responsibility for high traffic and congestion road system set up command centers of a sort, with lots of cameras, the ability to change signaling systems, put messages up on signs, maybe even open or close entrance gates and the like. In the center I worked at for a while (on their abandoned bespoke software), they also dispatched or coordinated with crews to do work.

Which brings up one issue, working for or with the government has its own special characteristics, and is going to be inevitable in this sort of career.

Good luck!

learc832• 12 years ago

One possibility is companies that make electronic devices for this

Thanks for the suggestion. My degree concentration is actually embedded systems, so that's interesting.

RhysU• 12 years ago

Traf-o-data.

peatmoss3• 12 years ago

Depending on what your interest in transportation engineering is, you may find yourself pretty disappointed. I can probably opine a little on this based on my background. I did primarily unix systems and security for the better part of a decade before getting a masters in urban planning, spending a few years working for a metropolitan planning organization including in their travel modeling group (these are usually the major public sector transportation modeling agencies due to federal requirements), and am now back in school working on a PhD in planning. I'm going to give you the traditionalist's view here, which may look a little drab / discouraging.

If you want to "fix traffic" I highly recommend running away now. You will not fix traffic as transportation modeler (I assume transportation modeling is what you mean by traffic simulation). The state of the art in travel modeling is not good enough that obvious solutions present themselves very often. Even when obvious solutions to some narrowly-constrained problem do present themselves, model results are usually one data point for consideration in a delicate dance between bureaucrats, politicians, and a general public with diverse and competing interests. You will also find yourself very poorly compensated compared to your peers who graduated in CS.

Another thing to keep in mind walking into anything transportation related (private or public sector), is that you are going to have a big ramp-up on domain knowledge / will start at the bottom of a low-paying totem pole. In my group in a public agency, we had a handful of people with PhDs, lots with masters, and I'm struggling to think of one person with a bachelors. And to be clear, a masters degree in civil engineering, planning, or public affairs qualified you for an entry level (assistant) position or maybe an associate level if you had some decent experience on top. I don't bring this up because I'm under the delusion that degrees make the person, but do beware that you're likely to run up against some biases if not hard degree requirements. In terms of domain knowledge, the transportation research board (TRB) is one of the national academies of science. I'm not sure the best scholarship always happens in this domain, but I bring this up to say that there's lots to learn... Also, transportation modeling is a very stats-heavy domain. Getting into the guts of transportation models is likely to require more than a passing familiarity with graduate level statistical methods.

In the public sector, the MPO I worked for was actually developing its own next generation travel modeling package, but opportunities in the public sector for that are as rare. The public sector has been running so-called "trip-based models" for decades. The software for these has been an off-the-shelf affair for some time, with actual model runs set up through either through monotonous clickety-click or ghastly macro language by intern-level positions guided by more senior modeling staff. Activity-based models are the next generation, and are being tested out by more ambitious agencies now. Once the benefits of these new models are proven out, smaller MPOs and city planning offices will likely adopt them. However, I don't see the the second wave hiring in-house coders--they'll wait until they can get an off-the-shelf package (or if there is justice in the universe, some usable open source version). I mention this because, despite the breakneck pace at which academia produces new and novel statistical and modeling techniques, consumers of these models are crazy slow at adopting new model types. This obviously has implications for what kinds of coding jobs exist.

You might find an in as a coder for a company that writes one of the major existing travel modeling software packages, but my understanding talking to them, is that they tend to be fairly small operations and have coders who also tend to have some domain expertise in travel modeling. I could imagine them hiring a straight-up coder, but breaking into the subject matter might be hard, as you could very easily end up doing pretty generic coding tasks with little relation to traffic engineering.

Reading through this, I'm realizing I sound pretty discouraging. If you constrain yourself to transportation modeling, which is what I assume you meant by traffic simulation, you may find yourself disappointed by the opportunities. This is the boring traditional world of transportation modeling. If you're willing to look outside transportation modeling, you may find some more interesting transportation-related coding jobs that are friendlier to people with more of a coding / entrepreneurial bent. Particularly in the startup space. I think there's still plenty of greenfield left for people wanting to sensor-up the city, ferry those data around, and then make sense of it all.

Qworg• 12 years ago

While I'm not the OP, I would like to thank you for such a detailed and knowledgable response! Sometimes, HN amazes me. :)

peatmoss3• 12 years ago

Thank you! Also happy to answer any specific questions you, the OP, or anyone else has. I'd meant to bang out a quick reply, but my response ended up ballooning a little awkwardly.

learc832• 12 years ago

Thank you very much for the in-depth reply. I was hoping someone in hacker news would know enough about the subject to give me some ideas.

As far as my interest goes it's nothing specific, more of a general interest. I've been interested in traffic patterns since my first retail job years ago where I noticed that customers tended to enter the store in groups causing long lines. When I paid closer attention I noticed that cars tended to arrive in the parking lot in groups and I assumed that traffic lights and stop signs were causing this. This got me thinking about how much traffic engineering affects every day life in ways other than just commute times, and I find it fascinating.

Just curious--does a masters in urban planning require a specific undergrad degree or would it be open to someone with a CS degree?

peatmoss3• 12 years ago

My pleasure. I think a lot of people end up doing transportation-related things as the result of this kind of curiosity. In my case, I got serious about bicycle and pedestrian advocacy as an undergrad after a beloved professor was killed crossing a street by hit-and-run driver. It took me nearly a decade after that to decide to do a masters in planning. Then the research bug bit!

Planning is belligerently multi-disciplinary. Provided you can spin a good yarn in your entrance statements, and your CS GPA wasn't in the toilet, having a CS degree is a fine asset.

One thing to think about (beyond "do I want to enter a field in which I will earn less than what I'd make with a CS degree alone?") is how you want to come at transportation. For example, sometimes people ask what the difference between a planner and a civil engineer is. I usually reply something along the lines of "if you want you build a bridge, call a civil engineer; if you want to know where to build a bridge, call a planner."

Planning, civil engineering, and public affairs all have their fingers in the transportation pie. If you're a planner, a masters program will try to prepare you for a pretty broad range of tasks that include knowing something about your problem domain, how to communicate with political types, and how to communicate to angry citizens with pitchforks.

Civil engineering tends to focus more on nuts and bolts kinds of problems in transportation. Oddly, I feel like civil engineering is more solidly represented in statistical modeling functions within planning agencies. I.e. civil engineers tend to be more quantitative. That said, I always had the option of taking courses cross-listed with statistics--the same coursework that is now making my brain bleed in the PhD program.

Planning tends to be more systems-oriented at the technical side, and more human-oriented at the drum circle side. The differences between public affairs and urban planning can get pretty minor at times--my masters program offered concurrent MUP with public affairs (and maybe civil?) options. I knew a number of students who did public affairs and planning.

My feeling (perhaps erroneous) is that planning masters students get a little more latitude to play than do civil engineering masters students. There are planners that do lots of different kinds of stuff--even just within transportation planning, never mind land use (more important than transportation), electric, sewer, water, etc. As a result of needing to do lots of stuff, you get more freedom to scratch your own itch. That said, getting a job with your masters in civil and PE is likely a bit easier than as a planner. But, you've already got a job-friendly degree ;-) For the record, I had an easy time getting work as a planner after my MUP, but I did have classmates who took a little longer to find relevant work.

My recommendation before going and upending your life for another degree would be to go read some good mass-market books that explore basic concepts and contemporary issues in transportation planning. In order, I'd suggest Traffic by Vanderbilt, The High Cost of Free Parking by Donald Shoup, and Human Transit by Jarrett Walker (disclaimer, I had a minor role in reviewing the latter). All three are written with a minimum of jargon, and all three will give you a flavor for the kinds of things planners think about. You might then find out if your city has some sort of group that meets to hack for civic purposes. These are cheap ways of scratching the itch...

Ask HN: Are There any Interesting Jobs in Traffic Engineering for a CS Major?

Related Stories