Caltrans e-toll problems & Internet stir

Ron Slade, an engineer at Caltrans created a stir by posting an Internet suggesting the Omaha NE systems integrator, MFS NetworkTechnologies had flunked out on a contract to develop electronic tolling for California’s 9 toll bridges. MFS was about to get a letter from Caltrans, Slade wrote, saying that their system had failed tests. There were “numerous flaws” in the equipment and he suggested that the contractor was in danger of having its contract ended.

We started checking out the story and we’d hardly gotten the incendiary Slade memo off the Internet and pasted it into our harddrive, when the Internet web site came up with a note saying that Slade’s item has been pulled. Attorneys representing MFS said it contained ‘several false, inaccurate and defamatory statements directed toward MFS and its participation in the construction of the (Caltrans ETC) system.’ "

Slade, although he drew attention to some real problems does seem to have got it badly wrong in places. MFS was not about to receive a letter of failure, according to Charles Price, the project manager for electronic tolling at Caltrans and the guy who would have to sign off on any such letter, and there was never an issue of MFS’ contract being ended. Pro forma the original contract ended June 30, as Slade wrote, but this was a coincidence and the contract also provided for discretionary extensions by Caltrans, and such extensions were always intended. Slade was overdramatizing, or at least misunderstanding, the situation.

What Caltrans project head Charles Price and MFS agree on is that the project is an exceedingly difficult one and that MFS has not yet succeeded in meeting Caltrans ETC requirements. To the extent the Slade thing on the Internet signalled that difficulty, it was accurate. Price characterized the Slade report as partly a “glass half empty or half full” issue of interpretation. Whereas Slade was portraying success-not-yet-achieved as “failure” Price and Caltrans’ contractor MFS both describe the same result as development work that is ongoing.

“We have got more work to do. We are breaking new ground with this system and it is tough, but we are confident we can do it, and we are certainly not failing. On the contrary this system is already exeeding state of the practice performance,” says MFS vice-president Bill Brogdon.

The major problem is an automated vehicle classification system MFS is developing for Caltrans. Now most automatic vehicle classification systems sort vehicles into 4 or 6 or so vehicle types, or else they weigh them. And most vehicle classification systems can cross-check with motor vehicle registry file of vehicle types or with the transponder account. But the legislated toll schedule under which Caltrans operates has 17 different types, with some of the distinctions being quite esoteric, and exceedingly difficult for machinery to classify to a high degree of accuracy.

In California there is a major concern about privacy. Caltrans feels the vehicle classification cannot be cross checking with vehicle registry files so it is shooting to have the vehicle classification system perform as a standalone.

MFS has not yet met this very difficult technical challenge, its representatives readily acknowledge. But they are working up some new configurations of the technology (laser diodes, light curtains and axle counters) and say they are very confident they can improve performance the few tenths of a percent needed to meet Caltrans demanding requirements.

Sheep 2c/head

Charles Price of Caltrans says that the legislation under which they operate has produced a far more complicated classification regimen than they would ideally operate. But it is their job to try and make an automated system fit the complexities of their inherited law-driven system of classifications and charges. At the Carquinez bridge some years ago there was a legislated toll rate per sheep driven through the toll gates (2 cents/sheep), an indication of how the toll schedule accumulated over time.

Because of the privacy issue Caltrans wants anonymous and transferable e-tags (a guy can use the same tag on his delivery van one day and that evening give it to his wife to use in her car), Caltrans have no vehicle data from the tag to crosscheck against their automatic vehicle classification. The integrated ETC system being tested at Carquinez crashed several times as Slade’s report stated, the Caltrans project director says. But, he says, Slade did not appreciate the fact that this was the result of a fixable problem. They had not yet installed routines to clear out old data, so the systems were crashing because they were overloaded. Once routines are established for purging unneeded data in a timely fashion the problem will be overcome. No big deal, says Price.

The Carquinez bridge has been fully instrumented for ETC for several months now and Caltrans has been testing the classification system on lines of vehicles at the manual collection lanes. 200 vehicles have been carrying e-tags.

The bottom line seems to be that Caltrans and its contractor MFS are having more trouble than they expected getting a challenging electronic tolling system up and running, and a startup is delayed. Caltrans is having further discussions with MFS about the scope of work. Caltrans’ Charles Price says: “We are delayed a bit. It has been difficult. But we are quite close to having an operational system, but we are not there yet. I’m hopeful we will have a system working for the public by the second quarter of next year. That’s my goal.”

Once when the system has been proved at the Carquinez bridge Caltrans plans to fit four more northern Bay area bridges and once they are complete do the final four in its 9 bridge system. (Contact Caltrans Jeff Weiss 510 286 4444)