This was cool (if a bit late): store locations for Trader Joe's, plotted against Google Maps. Last time I checked, the store list was just an annoying 3-column PDF file online that required a lot of scrolling to navigate.
Here's to hoping for more of these, having more stores geoplotting their location information, even if it's limited to pins on a map. It's a start after all. But that just leads to plotting multiple sets store chains on a map, and plotting an optimized route between them. Let's just face it: unless you live in Vermont, everything you consume comes from one of a finite pool of chain stores. The task becomes one of finding the most efficient way to bounce between them before moving on with life.
Ah, with the family away in Japan for most of the rest of the summer, it's time to put together my dissertation proposal. I'm (re)aiming for the end of August.
Monday, June 30, 2008
Thursday, June 12, 2008
Carpool talk
So I'm going through a bunch of articles about commuting choice, though I'm finding a lot of stuff about carpools this time. I think I just had my blinders on, focusing on just the bus. It's an asymmetric model- they provide the bus, and we hop on (if it suits us). Doesn't help much that even when the bus is full, they lose money on every run.
So carpooling suddenly seems interesting- within rigid constraints.
On the face of it, it's an easy problem. Who wants to go, when, starting city, ending city. Without any understanding of geography, the sites I've seen so far aren't so good at more flexible searches-- like what happens when a potential rider is somewhere in the middle of start/finish. But a little GIS in the mix can solve that part.
No- the real problem is that there's a large number of sites (I quickly found a dozen or so, but AFAIK they may number in the hundreds when you factor in on-line college "ride boards", Craigslist, etc.
What we really need is a standardized way for these lists to be searchable and interchangeable. At the end of the day, nobody cares which site the ride match came from- merely that one is found. There are a bunch of issues to care about-- notably privacy, safety, and reliability. But on the other hand, hitching a ride to Alaska seems a lot more interesting these days than sweating it out in Albany reading psych articles about trust formation.
So carpooling suddenly seems interesting- within rigid constraints.
On the face of it, it's an easy problem. Who wants to go, when, starting city, ending city. Without any understanding of geography, the sites I've seen so far aren't so good at more flexible searches-- like what happens when a potential rider is somewhere in the middle of start/finish. But a little GIS in the mix can solve that part.
No- the real problem is that there's a large number of sites (I quickly found a dozen or so, but AFAIK they may number in the hundreds when you factor in on-line college "ride boards", Craigslist, etc.
What we really need is a standardized way for these lists to be searchable and interchangeable. At the end of the day, nobody cares which site the ride match came from- merely that one is found. There are a bunch of issues to care about-- notably privacy, safety, and reliability. But on the other hand, hitching a ride to Alaska seems a lot more interesting these days than sweating it out in Albany reading psych articles about trust formation.
Thursday, June 05, 2008
You say carpool, I say dynamic ridesharing...
So, I'm (reluctantly) reaching an inescapable conclusion: the bus doesn't work for most of my target population. But neither does driving, for very long. So what's left? Carpooling, except nobody wants to carpool, either.
The news is running occasional bits about the bus-- not that you should merely ride it (people are now, up 10-20%) but now they're overcrowded at times. So why can't they expand service? Because the bus company can't afford it. Even with every seat full, they lose money on every run. When the rates were set a few years ago (I think it was sometime in the 90's, but I don't remember), diesel was about a buck a gallon. Now it's over $5.
Despite the hysteria about fuel costs (it's not the percentage increase so much as the rate that it's eating away everyone's discretionary income), what also hurts public transit service is that everything has gotten more expensive-- labor rates, insurance premiums -- while at the same time, public funding has been problematic. Somehow telling them vaguely to be "more efficient" hasn't been helping much, either, when by law and contract, they can't really change much.
So here's my service population: community college students. It's a thinly scattered population commuting to suburban or rural locations. The classic worst case for the provision of public transport. They drop out for a variety of reasons, mostly financial. Commuting is expensive, and many students drive to campus for 20-30 miles each way, 4-5 times a week, so we'll call that 150 miles a week, best case, times 15 weeks a semester. Assuming 25 mpg (averaging the kinds of cars I see in the lots, best case) we're looking at about $24 in gas per week, or $360 per semester, in the ballpark of the total costs of the books they're supposed to be buying.
Is $24 per week a lot? But that's just the school trip. Total gas costs if you live in rural areas are going to be high. Once your discretionary income is completely eaten up by rising food and fuel costs, that extra $24 just might make it hard to fill the tank to make it to class-- if you're working and can't fill the tank until Payday, that might just prevent someone from going to class. Then they fall behind. Then it might just go downhill from there.
And community colleges are notoriously non-social. Many students have trouble forming communities-- social networks. It's like a job in a way-- commute long distances, co-locate in rooms for a short period of time, then go home. Though on the job, you co-locate in the same room for more than 50 minutes, more than 40 days, before parting ways possibly forever, so there are even fewer opportunities to socialize, unlike colleges with student housing, in contrast.
If you don't know anyone, it's hard to do the kinds of things you'd associate with college-- meeting a diverse group of people, finding people to work with on homework assignments, and yes, finding the unlikely case of someone who lives sort of near you, with a similar class schedule, with whom you can take turns driving to school with. So in the age of Google Maps and Facebook, there might just be a technology solution to all that.
The news is running occasional bits about the bus-- not that you should merely ride it (people are now, up 10-20%) but now they're overcrowded at times. So why can't they expand service? Because the bus company can't afford it. Even with every seat full, they lose money on every run. When the rates were set a few years ago (I think it was sometime in the 90's, but I don't remember), diesel was about a buck a gallon. Now it's over $5.
Despite the hysteria about fuel costs (it's not the percentage increase so much as the rate that it's eating away everyone's discretionary income), what also hurts public transit service is that everything has gotten more expensive-- labor rates, insurance premiums -- while at the same time, public funding has been problematic. Somehow telling them vaguely to be "more efficient" hasn't been helping much, either, when by law and contract, they can't really change much.
So here's my service population: community college students. It's a thinly scattered population commuting to suburban or rural locations. The classic worst case for the provision of public transport. They drop out for a variety of reasons, mostly financial. Commuting is expensive, and many students drive to campus for 20-30 miles each way, 4-5 times a week, so we'll call that 150 miles a week, best case, times 15 weeks a semester. Assuming 25 mpg (averaging the kinds of cars I see in the lots, best case) we're looking at about $24 in gas per week, or $360 per semester, in the ballpark of the total costs of the books they're supposed to be buying.
Is $24 per week a lot? But that's just the school trip. Total gas costs if you live in rural areas are going to be high. Once your discretionary income is completely eaten up by rising food and fuel costs, that extra $24 just might make it hard to fill the tank to make it to class-- if you're working and can't fill the tank until Payday, that might just prevent someone from going to class. Then they fall behind. Then it might just go downhill from there.
And community colleges are notoriously non-social. Many students have trouble forming communities-- social networks. It's like a job in a way-- commute long distances, co-locate in rooms for a short period of time, then go home. Though on the job, you co-locate in the same room for more than 50 minutes, more than 40 days, before parting ways possibly forever, so there are even fewer opportunities to socialize, unlike colleges with student housing, in contrast.
If you don't know anyone, it's hard to do the kinds of things you'd associate with college-- meeting a diverse group of people, finding people to work with on homework assignments, and yes, finding the unlikely case of someone who lives sort of near you, with a similar class schedule, with whom you can take turns driving to school with. So in the age of Google Maps and Facebook, there might just be a technology solution to all that.
Subscribe to:
Posts (Atom)