Texas Transp Institute's two weighty reports on network tolls

January 14, 2010

Skymeter's Bern Grush calls it TDP pricing - standing for charges that involve measures of Time, Distance, Place. Others call them vehicle-miles fees or mileage charges. Sometimes they become VMT charges for Vehicle Miles Traveled. Others again call them road user charges or road use charges, hence RUCs. Texas Transportation Institute scholars call them Mileage Based User Fees, MBUFs.

Everyone has much the same thing in mind - generalized tolls, but it is a measure of the so far unfocused state of the discussion of generalized or network road pricing that there are so many different terms for essentially the same concept.

Grush's TDP, it seems to us, captures with an engineer's precision what we're trying to offer here - charges for road use that be adjusted for time of day, place of travel and distance traveled. Except it leaves out C for vehicle Class (car, truck, tractor-trailer etc.) It should be TDPC pricing.

TDP is an already a heavily used acronym - for thermal design power, thermal depolymerization, Texas Democratic Party, and a major political party in India that we're sure to misspell if we try to type it out. And TDP or TDPC is most uncatchy, just not memorable.

Texas Transportation Institute's Mileage Based User fees or MBUFs at least make a funny sounding word: 'em-buffs'.

But they aren't - as described by TTI - fees just based on mileage but on Grush's T and P and also the C he forgot.

In the past we liked 'road use charges' best, but most people perversely let an R slip in and say 'road useR charges.'

Any charge on a road useR such as a gasoline tax, a registration fee, a driver's license fee or a sales tax on a vehicle is a charge on a user of a road, and hence a road useR charge. The term is way too expansive to capture the notion of a road use charge.

RUCs need to convey the sense they are charges for use of the road, not charges on road useRs.

The distinction is too subtle for many people, it's obvious from the casual way they clip from road use to road user charge.

What the scholars at Texas Transportation Institute call MBUFs are network tolls because they envisage variable charges based not just on miles traveled - distance - but the RUC concept of tolls that can vary also by time and place and vehicle class.

A MBUF or a TDPC or a RUC is just a road toll really, albeit a much more ubiquitous one.

What we are grasping for is the notion of tolls over much or all of the road network rather than just the traditional notion of a toll on the discrete bridge, tunnel or highway.

We could combine the terms 'toll' which everyone knows, and 'network.'

Network tolling or net-tolls

We could then call them network tolls or 'net-tolls' to indicate they are tolls on a whole network of roads and crossings rather than just on a toll on individual roads or bridges.

TTI authors outline the deficiencies of the fuel tax based pricing:

- increasing fuel efficiency of vehicles drives down fuel consumption and perversely reduces revenues for roads

- 'alternative' powered vehicles (hybrid gasoline-electrics, mains-charged electrics) will be an increasing segment evading a charge for road upkeep

- fuel based charges can't provide for much needed variable pricing of crowded urban roads by time of day or level of spare capacity (congestion pricing)

Net-tolls are needed to send market signals to road service providers as to the value motorists place on improved service and extra capacity, and therefore how much spending on new capacity is justified - as well as simply to manage existing capacity for better service.

TTI authors Ginger Goodin, Richard T Baker and Lindsay Taylor sum it up (we substitute our term network tolls for their MBUFs):

"(Network tolls) are a desirable replacement to the fuel tax precisely because they address, or can be structured to address, the three main criticisms of the fuel tax.

"Because they are based on miles driven and not the amount of fuel purchased and consumed, their revenue base will not be threatened by the increasing fuel efficiency of the domestic automotive fleet. Depending on administrative structures for fee calculation and collection, they can capture miles driven by vehicles falling outside of the fuel tax collection framework.

"Furthermore, pilot projects by ODOT and PSRC have shown that congestion pricing elements can be applied so as to better allocate roadway usage among drivers when desired."

Their two reports are a careful and systematic study done on the issues to be decided in any move toward net-tolls, perhaps the most detailed discussion of alternatives yet produced.

The TTI authors see such road pricing as something that will evolve - implemented in bits and pieces, going through stages of development, rather than in simple sequence being planned, designed, legislated, and applied.

Too many institutions and parties have an interest in network tolling for it to come into being wholly formed.

Top down planning's fiascos

They cite the experience of the Real ID Act passed by the US Congress in 2005 as a warning against a federally designed and imposed program.  The TTI authors are restrained in their prose but the effort of the Feds to mandate national standards for issue of drivers' licenses for improved security ID has been an almost complete fiasco deserving of tabloid type hyperbole.

Where's Ray LaHood's rallying call: "A house divided against itself cannot stand. I believe this nation cannot endure, permanently, half real ID and half unreal ID..."

In point of fact this effort in federal direction has united the states! Against the federal government!

Not a single state was in compliance at the end of 2008 after the three years the US law provided for its implementation.

Weakly the Feds have simply provided "extensions" to avoid an admission they can't enforce the program they legislated.

Worst of all about half the states have said explicitly they reject the program, and won't go along with it.

23 actually have legislation prohibiting compliance with the Real ID Act.

The National Governors Association has produced a damning report saying the Act is vague on key definitions so won't achieve its purposes, and that it can't be implemented as laid down.

In contrast...

In contrast a program for sharing revenues between states and Canadian provinces - the International Fuel Tax Agreement - started in the 1980s at the state and provincial level developed into a working arrangement.

Federal support consisted of a later framework for reporting with certain key definitions for uniformity, but provided flexibility on implementation.

IFTA worked because of a consensus developed that diesel fuel revenues should be shared based on mileage traveled. The concept of a base jurisdiction - where the trucks operated out of - worked to reduce trucker compliance costs and simplified auditing.

A starting point for the TTI teams is that network tolls will be launched as something of a replacement for fuel (gasoline and diesel) taxes as the principal source of revenue and financing mechanism for roads. That suggests they could be levied by the various state governments and the federal government since they presently levy the fuel taxes.

But hopefully it would allow clarification of responsibility and less by way of mandates and matching funds, and each level of government trying to have a say and claim credit on every level of road above the local subdivision.

A key bonus is that network tolls could additionally, or alternatively, be applied by local government jurisdictions - cities, counties, joint powers authorities, or metro areas. Or by franchised private sector road service providers.

With the implementation of network tolling, other road user charges (all fuel taxes, registration, license fees) could be phased out within those jurisdictions as a funding source)  and the new road use charges used to direct and fund road operations and improvements in that area. That was a recommendation of one of two recent commission's into the future of transportation finances (NSTIFC).

The TTI authors point to public surveys that show the "bridge to nowhere" perception of politically directed public financing has come to characterize the current system of handout programs based on fuel taxes. This offers hope that a move toward a more market-dictated system centered on pricing will prove politically viable. It would clearly more efficient.

Privacy concerns can be addressed in many different ways - for example by doing computing and storing travel data within the vehicle - but there are some inevitable trade-offs between privacy and auditing and the ability to challenge perceived mischarging. If there isn't any record kept of movement the driver can hardly challenge the tolls levied.

TTI argue that an immediately mandated system of network tolls on all vehicles is the least viable option because of the retrofitting required.

It should start with trucks they say. And it could be extended to new vehicles with factory built-in units. The program could be voluntary with the incentive being the discontinuance of existing user charges. But they also argue against "balkanization" and suggest rather vaguely that it should be a federal role to set guidelines to prevent this.

Their report endorses the Oregon DOT plan as one the best trials so far for a transition to network tolling.

The institutional report of 69 pages:


Technology report inconclusive

The technology report is necessarily inconclusive, based on the fact that policy objectives will determine what level of technology is needed. The simplest alternative to the fuel tax, designed to avoid fuel efficiency's erosion of the fuel tax need would be an odometer-based charge with a flat-rate per-mile toll. Bit it would fail to allow us to do all the other things we want network tolls to do such as help manage congestion, and set prices according to market need.

It suggests the possibility of parallel technology options ranging between odometer-reads and fancy vehicle-based on-boarrd units (OBUs) in which motorists would have options so they could trade-off - auditability against the greater privacy of flat per-mile tolls, or deploy the more sophisticated system and take advantage of the greater pricing sophistication and accountability.

All the network toll technology breaks down into:

- roadway use measurement

- computation of the toll

- vehicle to backoffice communications

TTI doesn't go along with the notion that network tolling has to be GPS-based (satellite location finding).

A databus (OBD port) on all vehicles younger than 13 years provides speed-based distance measurement that is more accurate and less easily tampered with than an odometer and is designed to be accessible electronically.  A University of Minnesota team came up with this in a study last summer.

see http://www.tollroadsnews.com/node/4256

At the next level up this could be supplemented with radio beacon-based location stamping. A loose network of wireless beacons - mobile/cellular telephone towers could serve the purpose -  would not provide exact positional data but it would allow metro areas to be distinguished from countryside around or states and counties from one another in the distance data.

A tighter network would make use of both OBD speed and distance data an exploit a tight ubiquitous wireless network for very accurate positioning. Each vehicle would have a transponder (OBU) constantly searching for the nearest beacon and a ubiquitous network of available RF points - for example cell phone towers -  for the OBUs to work off.  The cell phone network would provide dense almost complete coverage and a "good degree of charging reliability," say the TTI authors.

Next step up the technology ladder TTI people see is the 'birds' (satellites) - GPS or GNSS. They get a little into the accuracy and reliability shortcomings of satellite location data:

"(M)easuring distances traveled based on triangulated location coordinates can result in inaccuracies. There are, however, corrective algorithms that ensure that even if location coordinates are not accurately identified, distances driven are measured within a tolerable range of accuracy.

"Reliability may also be an issue with wide?area communications?technology?based positioning. In some dense urban canyons or tunnels where signals do not penetrate, devices may go offline.

"Atmospheric disturbance can also cause long?range data transmission to fail. Again, there are solutions to this problem that involve imputing miles based on the data from before and after the signal loss."

Accuracy requirements and coverage needed will depend on the uses to which pricing is to be put. A flat rate per mile toll only requires distance data at the low end.  That's called a "thin client" approach.

At the high end congestion pricing of certain lanes in a broader highway requires far more accuracy and coverage.  That involves what they rather ineptly call a "thick client" not thick in the normal sense of dumb or stupid, but the opposite, a smart OBU that delivers a plentiful (hence thick) data stream. This system would deploy a data-heavy digital map and the related toll charges within the OBU. The "vehicle thick" approach could see the toll charges computed continuously from the map and pricing schedules within the vehicle.

This would be designed to satisfy the vocal constituency fearful of government tracking because only the finished toll charges due would be transmitted from the vehicle.

Another approach to privacy would separate functions with a third party intermediary transmitting the data.

An "anonymous loop-back proxy data transmission" is yet another ingenious, if rather complex, idea for providing confidentiality to motorists by separate back and forth transmissions, the first location data send being anonymous.

"OBUs would likely need to be designed with the ability for users to view their driving history, as billing entities are not
furnished with adequate information to provide a detailed billing statement. A final possible drawback to this architecture is an increased number of data transmissions."

Dense networks of readers with ranges of 100 to 1000ft (30m to 300m) could use wireless technologies including:

- dedicated DSRC in fact used at a range of 30ft (10m) as in present electronic toll collection

-  wireless local area networks (WLANs) chiefly WiFi

- Zigbee

The TTI authors see this extension of the current electronic tolling approach to networks as unlikely to be cost-effective because of the cost of covering so many low traffic density routes with readers.

Infrastructure costs they say  are "likely high (relative to other vehicle?to?back?office communication options) due to the need to build new infrastructure."

In contrast, they note, operating costs are relatively low: data transmission using localized technologies is "tantamount
to a file download, making it relatively cheap, regardless of amount of data sent; the primary operating
costs would be the maintenance and replacement of roadside readers."

"A major advantage of this architecture is its suitability to a variety of safety and traveler information related value?added applications, such as in?vehicle signing or warning messages."

They suggest it may be possible to greatly reduce the coverage of readers by use of 'mesh' networking - vehicle to vehicle transmissions.  

They also discuss a wide area constantly online design with a range between 300m and 40km (1000ft and 25 miles), probably using a wireless cellular system like GSM.

More complex still is a hybrid system of localized readers and the wide area continuously online.  

But they are most positive about this:

"A hybrid system would solve the system reliability and enforcement problems that exist with using solely detection?based communication, as data transmission would default to the wide?area connection if needed. Such a system would enjoy the flexibility of a wide?area, constantly online data transmission architecture.

"Perhaps most importantly, a hybrid system could reduce operating costs compared to a wide?area, constantly online architecture by itself.

"Because most vehicles would likely pass within a reader at some point during an uploading timeframe, the cheaper?to?operate localized data transmission technology could shoulder a large portion of data transmission.

"Moreover, because the wide?area, constantly online technology provides an assurance that data upload will occur, the network of roadside readers may not need to be as extensive; the ability to construct a leaner network of roadside readers could drive down infrastructure costs, compared to a stand?alone localized, detection based architecture.

"Finally, a hybrid architecture would enable the fullest menu of value?added applications."

5.9GHz with medium range transponder readers and vehicle to vehicle communications is seen as a possible longterm technology that could be adopted for pricing if driven by other applications. There they seem to have adopted the view of the E-ZPass intergaency group.

The technologies report of 44 pages:


COMMENT: These two reports are by far the best work yet done on what we've chosen to call network tolling. The one technology that seems to have been rather overlooked by the TTI team is the ISO 18000 6B/C series sticker tags - like TxTag or SunPass Mini - that come complete with high RFID read accuracy and writeback to memory. And, being backscatter they have no need for their own power. They'll work in tunnels, car parking garages, in among big buildings, anywhere the operator chooses to bung up a reader.

No GPS urban canyon problems here because you have the readers through the nooks and crannies of the canyons.

 And they are being sold now for under $2 apiece.  

Enforcement is going to be a major part of any pricing scheme, and that means cameras on poles or gantries to record a picture of the vehicle, its license plate and a time and date stamp.

On "privacy" civil liberties activists may protest cameras but their protests don't survive a commonsense test since we've all long since agreed to drive vehicles around with unique alphanumeric identifiers called license plates. There is nothing private about travel on a public road. There is no invasion of privacy in noting someone's presence on the public way.

We pay expensive police forces to fight crime and we want them to work efficiently - at least most of us do - so whatever the arguments we're likely to give them the technology tools they need, including the best cameras. Safety and national security issues - the Islamists - are also inexorably driving the development of networks of traffic cameras with license plate readers. The cameras have to communicate with back offices. Readers can be co-located, and they can piggyback off the camera comms.

All this discussion could be rendered moot by the incorporation of ISO 18000 6C sticker tags into vehicle registration and/or license plates themselves.  Electronic vehicle registration (EVR) using ISO 18000 6C sticker tags is already under way in Mexico and Brazil. All new vehicle registration stickers will containa  DSRC transponder.

The case for EVR is just as strong here in the US and network tolling could pick up on EVR.

TOLLROADSnews 2010-01-11

Further Reading

Leave a comment:

Upcoming Events