Drivers v. Robots, Part 7: Betrayal by Robots

Robots have decimated sectors of public transportation, like taxis and limousines. They have contributed to ridership decline in other sectors, like transit. And they have led to dramatic decreases in efficiency in still others, particularly paratransit and non-emergency medical transportation service. Robots will also contribute to a radical restructuring of public transportation as we know it (see “Drivers v. Robots, Parts 4 and 5” (NBT, November and December, 2019).

That robots have replaced large numbers of management personnel is well-known. This has been going on for decades. It is now clear that they can also replace drivers, and have already begun doing so. What is less well known is how they can trash the careers of top agency officials, company owners and industry leaders. This installment of “Drivers v. Robots” contains the story of a recent incident where robots ruined the career of a nationally-recognized transportation official in a major city. Names have been changed to protect the innocent (like in Dragnet) and the guilty.

Discord and Demise

It could have been anywhere. Let’s call it Gotham. Like in Batman.

In 2002, Gotham’s then-Director of Pupil Transportation asked me to attend a pre-bid conference. He was afraid no one would. The purpose of the project was to engage a team of consultants familiar with digital technology to recommend those to purchase. The thought of addressing the system’s problems with human beings was never considered. The central purpose was to eliminate them.

At the conference, I asked the  Director what his goals were. He could not identify any. When I asked if “safety” might be one, he agreed. When I asked if “security” might be one, he agreed. When I asked if “efficiency” might be one, he did not understand the question.

As the Q&A unfolded, we learned that the system had 80 human beings manning the phones after school hours to answer a single question, often expressed crudely and angrily: “Where is my f—— kid?” Department personnel could not answer this question. Many of the 6400 vehicles at that time were regularly behind schedule – even while schoolbus routes are usually identical from day to day during the entire school year.

Awash with scores of private contractors, Department staff did not know where any vehicle was at any given time, or which students were even on which vehicles. The contractors involved likely did. But the Department did not direct the passengers’ parents or guardians to call them.

The Department had no interest in hiring and training schedulers and dispatchers to keep track of this information. It was only interested in making the vehicles more on time so that the personnel who had been responding to inquiries could be eliminated.



More recently, parents and guardians were given an app by which they could obtain this information without bothering anyone in management. Of course, the app could provide no explanation about why the vehicle was late. And it could not provide this information to system management. But an app was a long way off in 2002.

Sometime after the pre-bid conference noted, a consultant was engaged. Eventually the Department acquired GPS equipment. One of its uses was to collect data to comply with Medicaid regulations, a program providing funding for schoolbus service, conditioned on the collection and submittal of GPS data (presumably to ensure that the funds were largely used to provide actual transportation). For a tiny fraction of the cost, far-more-meaningful data could have been collected by live Earthlings who would have become more familiar with the system and it’s issues through the effort. But this was not Medicaid policy. Nor was it Gotham policy. However, the digital collection approach was problematic: In 2009, not-so-mythical Gotham was forced to return the not-so-mythical sum of $535M to Medicaid for the City’s failure to collect the data required.

Partly to address this problem, Gotham hired a new Director of Transportation in 2011 — a nationally-recognized leader in pupil transportation and one of the nation’s experts in the transportation needs of special education students, exponentially more costly to transport than general education students. Two milestone’s place this Director’s accomplishments in perspective:

  • Shortly after his arrival, he went out to bid for contractors – the City’s first time in 37 years. Politically entrenched, most incumbent bidders retained their contracts. The major contractor, one of the nation’s largest, low-balled the bid so badly that its drivers walked off the job on New Year’s Eve. The contractor sold its fleet to another oligopoly at fire sale prices. Shortly after the Christmas holidays, Gotham’s new Director managed to keep the buses on the road.


  • Months after the Director’s hiring, an unlicensed motorist passed a schoolbus unloading, with its flashers ablaze and its stop arm extended, and mowed down an 11-year-old pedestrian crossing to her schoolbus. The bus stop was as perfectly-selected as reasonable, and the crossing procedures were articulated comprehensively. More interesting for purposes of this story, the bus serving the stop-in-question was relentlessly on time – often varying by not a minute from day to day. No staff were needed to man the phones for this route.

Beyond these accomplishments, the new Director was also asked to “fix” the GPS problems. Had he told the City it needed to hire and train project managers to monitor contractor performance, hire schedulers to design efficient routes, and engage assistants to review drivers logs, he would have been laughed out of the job interview. The best he could do was to joust at the digital windmills his drivers would not use. This buffoonery was of no interest to the robots. In this war of attrition and waste, the robots failed every task but won every battle.



As noted in installment #1 of this series (see “Drivers v. Robots, Part 1,” NBT, August, 2019), drivers traditionally trash GPS systems and robots. Eight years after hiring the new Director, a  September, 2019 study by the City’s Special Commissioner of Investigation revealed that 6000 of its 9000 GPS units were missing. How many were destroyed by drivers, compared to how many were never installed to begin with, was never discovered. It appeared to have never come up. Instead, the “Commission” scapegoated the Director and terminated his employment. His accomplishments were never on the ballot. Robots 1, Directors 0.

Robots and Ruin

As a footnote to goals which robots can achieve, the robots Gotham had purchased in the early 2000s (installed or otherwise) were not given the task of designing routes or selecting the stops. Instead, this task was delegated to Gotham’s contractors, with hours unconstrained by their operating contracts. The central directive was to get the buses to run on time, and as noted, to reduce (and eventually eliminate) the number of staff manning the phones. If the stop noted above was representative, no staff were needed.

On its way to providing the 20 minutes of service on the route segment encompassing the stop noted, the bus deadheaded a full hour to the beginning of the route, and an hour and 20 minutes back from the end of it. In other words, every minute of passenger time was matched by seven minutes of deadhead time. If one added in the time for pre-trip inspections (assuming they were conducted), every minute of passenger time would have been matched by nine minutes of non-passenger time. In a system with more than 7000 buses at that time, the robots contributed nothing to the control of billions of dollars in cost. Instead, they got a few score of personnel off the phones. The robots replaced a couple hundred meaningful jobs whose performance would likely have saved the City billions. The robots’ failure to collect data in the first decade of their alleged installation cost the City half a billion dollars in Federal funds. And if the route examined were representative, their use led to billions in waste. The failure of their alleged installation in the second decade, with 9000 buses by 2019, got the system’s Director fired. Just as it was of no interest in the first decade, this second decade’s multi-billion-dollar waste appears to have been of no interest.  

Footnotes, Fiascos and Fantasies

Particularly in light of the acknowledged costs of a financial fiasco eight years earlier, one can only ask why it took a transportation commission eight more years to discover the same or worse things still happening from its robots. One would need several more installments to merely list the idiotic findings. A tiny sample includes:

  • A single GPS unit cost $17,463.99.
  • The bid for them contained an incorrect unit cost.
  • The data they generated were not usable.



Even this tiny sample is deeply troubling:

  • One can purchase a brand-new, full size, rear-engine schoolbus, with all the bells and whistles, for what five of these units cost. In many economic environments, this sum would provide a dispatcher for a full year.
  • One can cite an incorrect total cost only if one cannot multiply, or worse, if one cannot read or type or, yet worse, if no one proofreads proposals, or yet worse, if the cost citation were deliberate.
  • Of course the data collected were not “usable.” To be usable, they must be examined by live Earthlings.

But these are only the footnotes that were not contradicted. Contradictions included:

  • Out of 6000 buses counted, and only 513 drivers turned the GPS units off (or did not use them). Yet the study found that 25% of the drivers logged into the system.
  • Since only about 25% of drivers actually logged into their devices, it is hard to know how 90% of the units communicated with one another.
  • Only $1.065M of the units were used. Yet if each one of the 6000 units (worth $17,463.99 apiece) computed to $105M, how is it that only $1.065M of them were used, yet only $7.535M of this expenditure was “a total waste”? This waste computed to only 8.2% of the first 6000 units. Where did the other 91.8% of their costs go?

One good thing about recycled garbage is that it is at least used. The same cannot be said about GPS equipment. One bad thing about GPS equipment is that the robots confuse those evaluating their performance so badly that they cannot perform simple arithmetic. In the future, when robots police themselves, this performance may improve. Of course, when robots police themselves, it is hard to know if such performance improvement will matter, even if it is more accurate.

Delusion, Denial, Deceit and Defeat

Beyond its preposterous mythology and worse arithmetic, the report’s conclusions noted that new GPS replacements (in 2019) would give the students’ parents “peace of mind.” It is hard to know how a commission which took nearly two decades to learn that hardly any of its drivers used the vehicle’s technology could predict that enough of them would do so in the future to justify this prediction. To be kind, such a prediction is a delusion. To anyone actually reading the report, it represents deceit, a pure cover-up. But this accusation may not be fair. Perhaps the most effective quality of a robot is its powers of persuasion. Otherwise, robots facilitate sloth and diminish accountability.



NBT readers may wonder if similar revelations about technology will catch up with motorcoach industry professionals. If USDOT was advocating for the replacement of paratransit service with driverless cars in 2015, what will it soon advocate about the motorcoach industry? Why would FMCSA want to continue monitoring the whereabouts of 33,000 coaches, with 4000 alone operated by owner operators? Replacing drivers with robots would solve lots of problems and eliminate lots of monitoring and enforcement jobs, particularly thousands of FMCSA field officers. The U.S. Jobs Elimination Program and reduction of the national debt could go hand in hand.

For now, the robots can only take over if we let them. If pretend mode continues, we soon will not have to let them. They will do it themselves because they will soon be in charge. For those less concerned about the long-term future, and only about their immediate futures, there is a lesson here. Robots will not simply take out your drivers. They will take out the top person in your organization. That person could be you.