Triboro's M Ascher - price was key in 1994 selection of Mark IV (FOLLOWUP)

Michael Ascher says the IAG did not capitulate to the special interests of the Garden State Parkway or any others in choosing Mark IV back in March 1994. He says cost was a major factor in Mark IV being given the sole source contract rather than the favorite going in, Amtech. Ascher was chairman of the executive committee meeting that decided the contract March 18 1994 in the Robert Moses Building of the Triborough Bridge and Tunnel Authority on a snowy Friday evening. In response to our report of a couple of days ago Ascher put his thoughts in an email we reproduce below.

Would NJ have walked?

Jim Crawford current executive director of the IAG who was moving into the position of chief executive of the South Jersey Transportation Authority around the time of the E-ZPass contract award told us in a telephone response to the article that the New Jersey executives of the SJTA and the Garden State Parkway would have walked out of E-ZPass if Amtech had been given the contract. He is not quite as sure that the NJ Turnpike Authority would have walked.

Mark IV did have tech migration

An engineer who was closely involved says that there was substance to the view that Mark IV offered a clearer path to technology improvements than Amtech. He says that Amtech was damaged by a candid statement by Mike Breslin of Amtech that their read-write offering was a technology "dead end." Mark IV by contrast not only had their own capable engineers but agreements with Hughes and AT&T for collaborative development.

Those technology improvements involved use of time division multiplexing and the slotted ALOHA protocols to handle multiple simultaneous messages and the essentially Hughes-designed ASTMv6 North American standard - too late to compete for the original E-ZPass - but adopted for truck regulation in the US and for tolls on 407ETR.

Mark IV did develop ASTMv6 transponders of this kind and also a dual mode Fusion transponder that will function as both an ASTMv6 and an IAG tag.

The Amtech read-write offering for the IAG was in essence the Texas Instruments inspired Type-21 open standard later made law in California, and adopted with small variations in Florida for SunPass. It was passive backscatter, responding only to incoming queries.

The virtues of an active transponder (Mark IV) versus passive backscatter (Amtech) were passionately argued, and remain an issue.

The engineer writes: "The IAG procurement required that any system proposed had to be commercially available on December ’92. Hughes’ system did not qualify so we were effectively locked out of the procurement. I had known Kelly Gravelle from way back to my SAIC days and his Vapor Canada days. The Hughes technology, like the Mark IV technology, was active (as opposed to backscatter) technologies. Hughes and Mark IV struck an agreement that, in part, said Hughes would supply a letter supporting the Mark IV proposal and Mark IV would identify the slotted aloha TDMA as the growth path for their technology to open road and more advanced technology."

Ascher suggested spreadsheet to address cost

Another participant in the 1994-03-18 meeting, like Michael Ascher underlines how important cost was in gaining eventual unanimity for Mark IV. He says they had a spreadsheet set up to show the costs of deploying Mark IV vs Amtech for each of the seven participating toll authorities. This cranked out different costs for different assumptions. The spreadsheet was available on a PC in a side room during the meeting.

Mark IV transponders were more expensive than Amtech's box, but per lane the Mark IV readers were cheaper. So overall the cost comparisons depended on assumed pieces of each purchased.

In most of the scenarios for most of the authorities Mark IV had an advantage, helping Mark IV to win. However transponders sales were in fact much larger than most of the scenarios assumed so looking back Amtech's cost disadvantage was less clear than it looked to decisionmakers at the time.

AT/Comm where lawyers outnumbered engineers

Ascher and others have emphasized their anger at the litigious antics of AT/Comm, a Massachusetts-based developer of a dual frequency active transponder system during this period. The company did sell systems to Maine Turnpike and Illinois Tollway and to a couple of tollers in Australia. They worked poorly - the lower write-back frequency was very liable to disruption by a range of other RF devices nearby.

AT/Comm systems were all closed down within a few years of installation. Instead of employing engineers to improve their product they employed a team of lawyers, and constantly filed nuisance law suits.

Toll authorities had to be extremely careful in what they wrote in internal memos and in any public comments for fear of AT/Comm's lawyers. The company for a while had a close link to IBTTA, the toll industry trade association creating difficulties in discussions there of electronic tolling.

Back office disappointment

Jim Crawford and others agree that one of the disappointments has been the inability of the group to have more joint back-office operations - handling customer accounts, and doing clearing house work. With all the focus on front end systems there wasn't enough energy put into coordinating the back office arrangements.

There was an RFP for a clearing house but it got a poor response and meanwhile the tollers were setting up state or individual customer service and accounts centers.

Everyone today is scathing about the early implementation of E-ZPass on the New Jersey Turnpike under the Whitman administration and commissioner Frank Wilson. New Jersey was saddled with a deeply troubled systems integrator that was a spinoff of WorldCom. And it unwisely adopted a financing model for electronic tolls based on supposed profits to be made from violations. No profit ever looked like being made, only losses.

This is what Ascher describes bluntly as "gimmicks."

Michael Ascher's comment in full:

"In its deliberation and selection of Mark IV, the executive committee did not cater to the special interest of (NJHA CEO Dave) Davis of the Garden State Parkway or any other member of the IAG. Cost was a major factor. At my suggestion we considered the total cost of the offerings to the region as an entity based on the total estimate of all transponders and the number of toll lanes to be equipped with roadside equipment. This approach placed the selection on a regional basis rather than the parochial interests of the individual agencies or any state. Mark IV's offer represented the lowest evaluated cost to the region.

"Maintenance of roadside equipment was not included in this assessment since some agencies planned to do this in-house. However, TBTA determined that its cost, including maintenance that was to be contracted out, favored Mark IV. The notion that cost is not important to agencies with high toll revenues is absurd.

"Shortly after the selection, the revolving door of New Jersey transportation officials led to an examination by New Jersey whether the technology selected may be out-dated. This was brought about by AT/Comm who saw this as an opportunity to get their foot in the door by tearing the fabric that had been created by the IAG.

"AT/Comm engaged a lawyer/lobbyist who was a former NJ transportation commissioner with close political ties to the Whitman administration and Frank Wilson. As a result Wilson engaged Booz-Allen to evaluate the chosen technology. After a year, (Booz-Allen) concluded that the technology was indeed current.

"AT/Comm then went on to Massachusetts and tried to humiliate transportation officials there with a barrage of attacks through ads in local newspapers. Soon after, they went out of business.

"Wilson then explored a number of gimmicks to place the cost of operating the NJ E-ZPass system on others. These initiatives proved to be disastrous. Walter Kristlibas was hired by the NJ Turnpike. after his retirement from the Port Authority of NY and NJ to clean up this mess.

"I suppose there will always be some that will want to re-write history, but the fact remains that even 14 years after that exhausting executive committee meeting at Randall's Island, E-ZPass remains a hallmark of success of regional cooperation, crossing difficult geographic, political and institutional borders."

Kennedy School study

When we wrote the report on the 1994 selection we weren't aware of the existence of a two part report on the origins of E-ZPass by Susan Rosegrant for the Kennedy School at Harvard. She interviewed many of the main actors in 2004. She says the driver for the IAG was the research finding that the average motorist trip in the New York/nNJ metro area involved the use of 2.4 toll facilities.

[This 2.4 number has to be way wrong. There are about 10.5m motorist trips/day in the NY/nNJ metro area and a lot less than 25m (10.5mx2.4) toll facility passes. The more likely figure is 5m tolled trips/10.5m = 0.5 toll facilities passed per trip. Maybe the average toll trip involves use of 2.4 different facilities, but that seems high too.]

Anyway, whatever the number, with four toll authorities in the New York metro area most motorists were likely to operate over several toll authorities facilities during the course of a year, and many on a daily basis.

That was the thought behind the need for interoperability.

Discussions of interoperability started by L Yermack 1989

From 1989 onward the New York toll agencies discussed the need for interoperability, but Rosegrant writes, there was little history of collaboration. In the spring of 1990 Larry Yermack, then chief financial officer at TBTA persuaded the chief executive of the parent MTA to propose the first meeting to the PANYNJ to coordinate electronic tolling. Tony Barber then at PANYNJ is quoted as saying the toll authorities were "very parochial, very independent, not terribly communicative one with the other, and didn't have a track record for sharing information or cooperating."

The first meeting of what came to be the IAG took place during an IBTTA meeting in New York in June 1990 and included TBTA, PANYNJ, NYSTA, NJHA, NJTA and the PTC. (No NJEA or SJTA yet).

1992-95 deployment first bruited

They produced a schedule for the first electronic toll operations by 1992 and a complete regional system by 1995. The decision was also amde then to be an informal cooperative rather than a legal entity - because authority lawyers said it might create conflicts with the powers of each authority, and because it might facilitate law suits. One agency one vote was decided then as the rule.

From the beginning Rosegrant says it was NY vs NJ with PA the swing vote, quoting Mike Zimmerman then of NYS Thruway.

In January 1991 work started on developing technical specifications for the electronic toll system preparatory to issuing an RFI.

99.95% accuracy was soon established as a standard.

Read-only or read-write?

By the fall of 1991, the toll authorities with ticket systems were insisting on read-write technology. For several months this was a major source of disagreement, the barrier system agencies concerned about the higher price and lack of experience with write-back.

In January 1992 the compromise was reached – a preference for read-write but read-only to be acceptable so long as there was a 'migration path' to read-write.

Carlos Nicot of TBTA worked fulltime on the procurement issues - how to evaluate proposals. Some agencies were against departing from a lowest bid award, but the argument for 'best value' eventually won the day.

The group also decided on a sole source supplier but the winner would have to agree to license a second source supplier - a provision that was later dropped.

The name E-ZPass was chosen by the marketing committee in November 1991. Linda Spock then of PANYNJ is generally credited with the term, and the PA has ownership of the brandname to this day.

A single back office for all the agencies was part of the plan in 1991 - but it was, of course, never achieved.

Violation enforcement required laws

State laws had to be passed to support enforcement of electronic tolling. Draft bills were written and lobbied for in the state legislatures.

The first technology deadline of April 1992 saw eight firms submit proposals. An evaluation committee worked for a week at a motel in Albany NY and reported that Amtech and Mark IV stood out above the others and should be shortlisted. They both proposed read-write systems.

In June 1992 testing began at Spring Valley on the Thruway and at the Hillside plaza on the Garden State Parkway. Hired drivers using rental cars did thousands of passes under ET equipment.

There were negotiations of several days with each company separately.

They came to nothing because both vendors failed the first round of tests. They had cost the IAG over $1m. Further test costs, the IAG decided, would be borne by the vendors.

Balkanization looming

Meanwhile the Thruway was breaking ranks because of the impatience of Peter Tufo chairman of the Thruway board. NYSTA put out its own RFP for a back office and began electronic tolling with read-only Amtech tags at the busy commuter toll plazas at Sping Valley, the Tappan Zee Bridge and seven other toll points. It did lease rather than buy the system from Amtech to underline its willingness to convert to the new technology when IAG got its act together.

The quick Thruway installation stirred up union hostility. The union successfully forced the authority to post 5mph speed limits through electronic toll lanes. But the technology worked well, upping lane thoughput to 650veh/hr vs 350 in cash lanes.

The second round of testing saw both vendors pass in December 1993. They achieved almost identical read rates in the tests. It was difficult to argue for one or the other on performance.

At this point the working assumption was that the IAG might top out an at eventual 1 million tags (there are now 16m and they haven't topped out).

One back office flop

In June 1993 the IAG put out an RFP for a customer service center or back office.

The Pennsylvania Turnpike broke with the other six, saying it would do its own.

And the IAG's RFP only brought two proposals, and only one was deemed 'responsive' - Lockheed. This Lockheed unit had recently been embroiled in a corruption scandal in New York City in which a councilman killed himself. TBTA and PANYNJ said they couldn't go along with Lockheed being selected because of the scandal.

So in November 1993 both proposers were formally rejected.

At the end of 1993 the new Republican administration in Trenton saw Frank Wilson appointed transportation commissioner. He intervened heavily demanding to know why read-write was being insisted upon. For a while it seemed as though the New Jersey tollers, at Wilson's insistence would do their own thing and implement the cheaper and proven read-only.

The New York Times and other newspapers pointed up the disadvantages of twin incompatible toll systems and Wilson backed off.

We're-here-to-help threat from FCC

A third potential problem arose when the FCC began hearings on providing a special new bandwidth for ITS, including electronic tolling. This threatened to be a major setback because all the existing technology was in the open 900MHz band. The IAG lobbied strenuously against the proposed reserved bandwidth, and that move was delayed sufficiently for the iAG to deploy in the unregulated 900MHz frequency.

The March 18 meeting was seen as crucial to the future of the IAG project because of the pressures of balkanization. Failure to reach the needed concensus about the front end procurement could be fatal to the whole effort to create one interoperable system.

Prices

Rosegrant writes that Amtech's transponder price was a third less than Amtech, and that both companies proposed readers costing around $13k, the big difference being that Mark IV's reader read eight lanes to Amtech's single lane reader.

Based on projections of 970k eventual usage of transponders Mark IV was clearly cheaper for the IAG as a whole but it was more expensive for TBTA and PANYNJ.

Rosegrant writes that the final vote for Mark IV was at about 8pm. She quotes Mike Zimmerman of the Thruwayas saying that his executive director John Schafer voted for Mark IV although the authority had decided on Amtech. Shafer was persuaded both would work equally well but they needed to keep New Jersey from leaving.

Ascher is quoted: "We trusted one another, we put the whole—the region—ahead of the individual and parochial interests of the agency, and we were able to cross these geographic and political borders."

To buy the Rosegrant report in two parts go to:

http://ksgcase.harvard.edu/casetitle.asp?caseNo=1818.0

http://ksgcase.harvard.edu/casetitle.asp?caseNo=1819.0

TOLLROADSnews 2008-03-31