2009/01/30

Parking meters: monuments to incompetence

From the San Diego Daily Transcript comes one of the best stories of the idiocy of municipal parking management I have ever read -- and I have read many. Surprisingly, the same state that brought you this sad tale also brought you the best book on the subject: Shoup’s High Cost of Free Parking. The text came to me via email and I cannot find an online source so here it is in its entirely.

Parking meters: monuments to incompetence
By LARRY STIRLING
January 21, 2009

While traipsing around historic Tallinn, the nice little capital city of Estonia, the major feature of which is a crenellated castle, the last thing I expected was to encounter an innovation in public administration. After all, what could that tiny country teach San Diego except how to endure repeated Russian military occupations?

Our small group had chipped in to obtain the services of a local guide who drove us around in her private car.

Each time she parked and we got out, she would walk to the curb to observe a number painted there and then punch it into her cell phone.
It turns out that the numbers constituted the Tallinn city parking-meter system. By recording her car location through her cell phone she was effectively dropping a coin into a virtual parking meter down at city hall.

There were no actual parking meters. They probably would not work well during the Tallinn winters. But, more importantly, there were no meters to buy, install, maintain, collect from, or replace as a result of old age, vandalism, or a bad parking approach.

The purpose of renting space on public streets is two-fold: the first is to create turnover among parkers so that incoming customers are more likely to find a place to park. The second is to raise money to support the city.

The Tallinn system accomplishes both objectives while the city of San Diego system accomplishes only the prior. The cost of installation, maintenance and servicing San Diego parking meters, including the employment of those hardy souls that walk around collecting the coins and then wrapping them defeats the revenue purpose.

Indeed, collecting parking revenues above the cost of operating San Diego parking meters would constitute a "tax" which must be approved by the voters and that has not happened. So parking meters are by law a break-even or, more likely, a money-losing proposition.

As a practical matter, the San Diego parking-meter program's primary purpose is to operate San Diego's parking-meter program. As such, it constitutes what one wag calls "a self-licking ice-cream cone."

If going digital is too big a mental leap for the city fathers of the home of Qualcomm, then, as an alternative to the grey-metal forest of obsolete parking meters, the city could adopt the Tel Aviv, Israel system of the "meter-less parking meter."

Tel Aviv, as so much of the world, cannot afford the puerile public-administration antics of San Diego.

Instead of feeding parking meters, drivers purchase city parking certificates through local convenience stores and gas stations.

When someone wishes to park, they grab one from their dash-board stash and then select and punch out appropriate chads indicating time and date. Then by use of a sticky strip, they attach the document to the inside of the driver-side window. This makes the document readily available to parking-enforcement officers.

The advantages are manifold. No meters of course. Revenue collection is done by local merchants who keep a small percentage for their trouble all the while enjoying increased foot traffic for their stores.

However, if our city fathers just cannot let go of parking meters, they could still save a huge amount by adopting the system used in Paris. There are typically only two parking meters for an entire block. Once you park, you walk to the end of the block, insert your coins or card, obtain a receipt and place it on you dashboard. The short walk is good for you and the city evades most of the costs that San Diego endures.

Each of these ideas has been presented to San Diego along with dozens of others for improving city operations, most, and probably all, of which have been ignored.

Instead our leaders seem to take pride in informing us that they have failed in their primary tasks of balancing the budget without further enslaving us through ever-increasing taxes.

We are told we are going to have to "share the pain." Really? What pain will they endure? Oh sure, they will reduce services to the public, but none of the big wigs are going to be laid off or take a pay cut.

In a previous column, I published a clear analysis of a multi-million dollar continuing cash opportunity for the city to help it balance its budget through taxi licensure. It's like "dynamic parking."

Just in case the city staff was too busy licking ice cream to read The Daily Transcript, I sent them personal letters including supporting material. So far officials have not deigned to respond. But they have announced that they want to charge fees for our trash pick up. What next, street fees, beach fees, fee fees?

When it comes to public administration, San Diego is the obsolete-parking-meter capital of the public-administration world.

Stirling, a former U.S. Army officer, has been elected to the San Diego City Council, state Assembly and state Senate. He also served as a municipal and superior court judge in San Diego. Send comments to larry.stirling@sddt.com. Comments may be published as Letters to the Editor.

2009/01/25

Insuring the Uninsured

Recent news in the UK reports that uninsured drivers cost law-abiding motorists £400 (US$557) million annually. Another report claims this costs compliant motorists an extra £30 (US$42) a year. The UK is proposing new legislation to make it an offense to own an uninsured vehicle, to say nothing of driving one. This cannot be so simple, as there are many reasons to own a vehicle but not drive it or to drive a vehicle that you do not own.

Perhaps there is another solution. Government-managed pay-as-you-drive (PAYD) insurance as a fall-back safety net can be imposed on every vehicle at the fuel pump. This can be set as a per-gallon premium, and immediately waived given proof of insurance.

Paying at the pump is not a new idea as a delivery mechanism for PAYD coverage. The problem is that such insurance is best scaled to a driver’s experience and driving record – i.e. PAYD is not simply a distance-based insurance, but a weighted, distance-based insurance. Hence paying by the gallon is very crude – although certainly more efficient than paying by the year. Furthermore paying at the pump would be expensive on a per-mile basis so would encourage purchase of more usual coverage or even further encourage the offering of PAYD policies that are properly metered by distance rather than gallons and scaled to the driver’s record.

Pay-at-the-pump should not be the normal payment system. Rather, I propose it as a simplifier for rental-car insurance, temporary insurance lapses, occasional drivers of antique vehicles, borrowers of a friend’s car, etc, as well as a fair way to make uninsured drivers insured.

The cost of adding the premium to fuel is small, but the cost of the payment-proof system and the associated labour of checking compliance before waiving the premium would not be trivial. The question is can it be done for much less than £30 per honest motorist per year?

I believe it can be. It would require that insurers provide each insured with an electronic ID card that holds the insurance expiry date, vehicle descriptor, plate number, etc. That part would cost far less than £30. A more significant part is the cost to the operators of pumping stations for compliance checks and the changeover costs for their cash-management systems. But if the pre-gallon premiums were set properly, the extra money in the system would leave the insured motorist whole while taking uninsured drivers off the road.

2009/01/10

Privacy, Location and Innovation

This entry in a Location-Based technology blog, relates to modifying data in a way that enables its public use while preserving privacy. [scan this entry to follow below]

This led to two comments by Bern Grush and Fosca Giannotti, respectively:

Location privacy re GPS tracking will be the single most critical issue for Time-Distance-Place (TDP) road-pricing (called VMT charging here in US and specifically called for in the new report from the [US] National Commission on Surface Transportation Infrastructure Financing.) The problem for that application and at least some solutions that address it are similar to Wood's view and solution. We patented something similar (NOT the same) in the recent past that allows us to "fuzzy-up" the entire journey to an arbitrary degree. BUT it was not filed under "privacy", rather as a method of non-line-of sight noise mitigation, trip descriptor reliability (for consistent charging) and compression. A year after it was granted we realized we also solved the privacy problem. I find that in addition to privacy (a concern I share with any person or motorist) and a knee-jerk response of "NO GPS" is that it stifles innovation WITHOUT providing real privacy. I wrote about this here. Bern Grush on 2009-01-09 12:18

The issue of Location Privacy is broadly studied within two major communities: Location Based Services and privacy preserving Data Mining. An extensive state of art may be found in the book: "Mobility, Data Mining and Privacy" edited by [Fosca Giannotti] and Dino Pedreschi. Such a book is one of the results of the European project GeoPKDD: other results are algorithms for anonymization of trajectories (trajectories are reconstructed by traces left behind for example by cellular phones). The idea of generalization is extensively used in anonymization. In this case what I see as critical the definition (and consequent labeling) of private and public as absolute concepts. Fosca Giannotti on 2009-01-09 16:53

The GeoPKDD site is a valuable resource for those concerned with Privacy and Innovation.

2009/01/02

GPS more private than E-ZPASS?

Rob writes:
“There has never been a successful defense against court orders to [surrender location data from tolling systems]. Most coming from divorce cases…”
I doubt this claim is true. But let’s assume for moment that it is. The kind of systems Rob refers to (since no better ones exist in North America) are the kind of electronic tolling system that have a $20 Automatic Vehicle Identifier (transponder) in your car and a million dollar reader-gantry-camera system above the road to read it. Such a system MUST retain your vehicle time of passage as proof to collect payment, UNLESS, as is offered by both E-ZPass and the “407” systems, you have a special pre-paid transponder that signals the gantry-reader that you have prepaid and your anonymous account is debited. And of the millions of E-ZPass and “407” drivers only a handful of drivers take advantage of this fully anonymous capability.

Surprisingly, it turns out that the opportunity to preserve your anonymity is GREATER with satellite tolling because the $100 device inside your vehicle can handle the entire transaction. There is no need for a million dollar reader-gantry that has to remember your passage! And yes, there is also the opportunity for greater invasion of privacy, but no country in the EU will permit this and I doubt the US will either. So far, every country that has addressed this has called for full privacy.

Road-metering innovation needs to balance two critical matters: the need for sustainable mobility and an entitlement to privacy.

Skymeter offers multiple levels of privacy (“dial-up privacy”) for our telemetric system from the fully opaque (level 0) where no data is forwarded or retained (much less available to others), to the fully transparent (level 4) where you can sell your data to marketers. There are three degrees of privacy in between, two of which, if subpoenaed would be surrendered and one of which can only be surrendered by you, unless you have already deleted it. It is also the case that we make it possible to drive and pay without using such a meter, but you will likely pay a monthly or annual fee that may be more expensive. And your plate will be imaged and stored more often, actually reducing your privacy. This last option and its cost is out of our hands, we are road-use-meter engineers, not policy makers. We solve the problem of reliable and private metering, we do not make the toll decision – or operate the fund collection and distribution systems. Road-owner-operators do the former and telcos do the latter. We enable everything in between.

The reason my company doesn't collect data in our level-0, anonymous-version services, is that we have built no way to retain the data once the trip is debited from your pre-paid account (which happens as soon as you are parked). So when you arrive at whatever place you should not have arrived at, the trip is debited and the journey erased; a message is sent to a billing operator to move the correct amount of money from a numbered account X (to which you had previously and anonymously deposited funds) to any number of road operator accounts corresponding to the road-owners whose roads you just used. When you arrive back home from whatever place you should not have been visiting the same thing happens. If you were arrested in your driveway, the trip data would already be gone, and so would the billing record. By the way if you were stopped along the way, the elapsed portion of your intended trip would be clipped, debited and deleted, as well. An easy way to clear your level 0 device is to stop for milk on the way home! The only problem is that you have no way to audit the bill.

Level 1 retains the location data for a numbered account X on your device and allows you to read it to check a bill. Once paid you may elect to delete it, or keep it for your records. It rolls off as memory fills. It holds about a year of normal driving.

Level 2 does the same but at a server where many additional services can be provided. Again the data is associated with a numbered account that only you have the password for. If you have reason to believe you might have trip information subpoenaed, select Level 1 if you need to self-audit – or Level 0 if you do not.

Level 3 is level 2 with permission for selected others to see your data. As an example, Bob, who is a problem to insure because of a record of driving violations may be insurable by a firm who examines Bob’s driving habits (speed, acceleration, time of day) and sets premiums accordingly. Clearly, this is an invasion of privacy, albeit one that Bob agrees to in order to be insured.

Level 4 data is sold to advertisers who send you coupons for opt-in things near where you park. This aspect is STILL anonymous (as in there is no need for your vehicle or personal ID), but anyone with a privacy concern might get antsy.

Now there is no way for Skymeter to know which levels any country would permit or disallow. We know of one (not the US) that will not allow level 0. And we know the EU is keen on ensuring only Levels 0 and 1.

So, the technology for absolute anonymity is possible, and it is ready. The issue for you is to be sure:
  1. Your government permits the privacy protection level you demand (they will be tolling your vehicle).
  2. Your government permits (or demands) an independent 3rd party to test that the levels behave as promised and that the more secure levels 0 and 1 cannot be compromised.
  3. If you choose anonymity, be sure to keep money in your account. (An empty meter turns off a signal that tells roadside cameras not to take an image of your license plate.
Ironically, having NO meter means your license plate is captured, having a working meter means it is not. So you should choose Level 0 or 1, rather than pass altogether.

Hello Hello Anybody Home?

It is really getting hard to believe that with all of of the criticism of the flat-rate London Congestion Charge – by advocates of congestion pricing, no less – and all of the talk of VMT-pricing [1] in the United States (which hopefully is TDP-pricing [2] by the wrong name), that intelligent people keep coming up with cockamamie flat rate schemes.

In San Francisco, 'congestion pricing' is something they're sneezing at.

California Dreaming: Even In San Francisco, People Don’t Want to Pay to Drive Downtown.

It flopped in NYC. I predict (and so do you!) that it will flop in San Francisco.

Charging a fixed fee is always economically inefficient. Whether you pay $6 or $20, once paid you have no incentive to drive less or differently. This is also unfair to those who do drive less. If Mary drives 13 miles in the zone, and Susan drives two, why should they both pay the same? For the same reason that all-you-can-eat diners gets fat, all-you-can-drive-cities get congested, we see that in London already. The London Congestion Charge may be well meaning, but it is wrong headed. Stop admiring it. The point has been made. Move on. Pay for exactly and only for what you use. Demand variable TDP-pricing in place of the gas-tax. Its fairer and more effective. And it can actually cost less and be anonymous (now those are nice surprises).

BUT some good news, here, Oregon seems to get it, and they're not all that far away from California. How come? What makes it worse is that tons of money were spent to discover what the Governor of Oregon now knows. Can somebody call someone in San Francisco please?

[1] Vehicle-Miles Traveled; [2] Time-Distance-Place