With automatic trains, why are train strikes in New Jersey still a thing?

In March 2016, the New Jersey Transit was held hostage by unions, representing over 1,000 train conductors and 400 engineers, to negotiate health insurance premiums in order to avert strikes in the region. With people as a key risk in New Jersey Transit’s operating model, should New Jersey Transit look to automated train lines as a solution?

In March 2016, the New Jersey Transit was held hostage by unions, representing over 1,000 train conductors and 400 engineers, to negotiate health insurance premiums in order to avert strikes in the region. With people as a key risk in New Jersey Transit’s operating model, should New Jersey Transit look to automated train lines as a solution?

In its purest form, automated train lines function without people and run entirely on systems of sensors and computers. The world’s first fully-automated train line was opened in Kobe, Japan, in 1981, and since then, developments in train technologies have yielded significant benefits. These benefits include reduced reliance on labor (and risk of strikes), improved safety, enhanced quality of service, and reduced waste and costs. Despite these benefits, New Jersey Transit and many other U.S. States have not implement driverless trains. Indeed, less than 5% of the U.S. train network is fully automated[1]. This blog discusses first the benefits, then the obstacles, faced by digitizing train systems in an attempt to answer the question posted above.

The first benefit of automated train lines is improved safety over traditional trains as they are not prone to human-error (by definition). According to the U.S. Department of Transport Federal Railroad Administration, 39% of accidents are due to human error.[2] Indeed, that according to officials, automated trains may have prevented the Metro-North derailment in 2013 that killed four people in New York when the conductor fell asleep [3].

Another benefit is that automated trains can increase capacity by up to 50%[4]. Train frequency can be increased as headway time (or time between each train) can be reduced to 80 seconds. Also, the setup time of trains can be reduced as train deployment can also be made fully automated. Further, the reduction of space needed for drivers can increase passenger capacity by ~5%.

In addition, automatic trains are more environmentally sustainable as they consume less energy. Acceleration, traction, and braking – processes that are controlled by human operators under a traditional system, can be optimized at each part of the journey, leading to greater energy efficiency. Energy savings are estimated at 30%[4] in some cases. Further, because train speeds are more accurately controlled, trains become more punctual and quality of service increases[5].

Also, with fewer people needed to operate trains, people would become less of a bottleneck in the train system, and the risk of train strikes would become less of a problem, as the New Jersey Transit recently (re)discovered. Furthermore, staffing requirements can be reduced by up to 70%, resulting in significant cost-savings in the long-run[6].

Requiring fewer people to operate train lines is also a key reason why implementing fully-automated systems is a challenge. Unions recognize that this technology will reduce their bargaining power, and hence are lobbying hard against this. We can see this dynamic play out in the push to automate London’s tube, a situation in which unions accuse administrators of gambling with public safety in order to pass populist ideas.

As alluded to above, public safety is also another concern. Many people believe that driverless trains are unsafe. This aversion is understandable, as after all, we trust another person’s interest for self-preservation to be more closely aligned with ours, than a computer’s. These safety concerns are sealed in the minds of the public though incidents such as the Lathen train collision in 2006 that killed 23 people in Germany. Consequently, decision makers have been fighting an uphill battle to receive public support for driverless trains.

Finally, capital investment is needed upfront as new infrastructure needs to be added to trains and stations. Whilst raising funds may be a challenge, these investments have shown to pay themselves off. Studies and surveys have shown that the investment can be recovered within a decade, and that returns are typically 10% to 15% (or even up to 30% in the case of the Paris Metro)[6],[7],[8].

Given the significant operational and financial benefits automatic train lines provide, I think it will be a missed opportunity for New Jersey Transit if it were to carry on its operations “business-as-usual” without using their encounter with the unions as a catalyst to initiate change.

(698 words)


[1] International Association of Public Transport, World Report On Metro Automation, July 2016, http://www.uitp.org/sites/default/files/cck-focus-papers-files/UITP_Statistic%20Brief_World%20Metro%20Automation%202016_Final02.pdf

[2] U.S. Department of Transport Federal Railroad Administration, Rail Safety Fact Sheet, https://www.fra.dot.gov/eLib/Details/L17342.

[3] NTSB wraps up on-site investigation of Metro-North derailment, cites need for PTC, December 11, 2013,  http://www.progressiverailroading.com/ptc/news/NTSB-wraps-up-onsite-investigation-of-MetroNorth-derailment-cites-need-for-PTC–38692′.

[4] Siemens AG Press Release, April 2012, Fact Sheet, “How does a driverless metro system work?”

[5] How do driverless trains work?, April 29, 2014, http://askascientist.co.uk/technology/driverless-trains-work/

[6] Cohen, et al., Impacts of Unattended Train Operations on Productivity and Efficiency in Metropolitan Railways, https://trid.trb.org/view.aspx?id=1337734

[7] Observatory Of Automated Metros, The Case For Conversion, http://metroautomation.org/the-case-for-conversion/, retrieved on November 16, 2016.

[8] Eric Jaffe, April 13, 2015, The Case for Driverless Trains, By the Numbers, http://www.citylab.com/tech/2015/04/the-case-for-driverless-trains-by-the-numbers/390408/.


Fitbit is the current market leader, but are its steps limited?


Thrown from the nest: Nest Lab’s Challenges and Opportunities

Student comments on With automatic trains, why are train strikes in New Jersey still a thing?

  1. I see many parallels to this problem in the debate over driverless cars. Many feel that driverless cars will revolutionize the transportation industry (myself included). By optimizing the traffic system, they will reduce accidents, relieve congestion, and optimize energy consumption through general efficiencies. Many of the same points were made in your post concerning automated trains. To your point, unions will never cease to exist to represent workers. However, I feel that it would be in the best interest of the unions to find new roles for their members, through training to fill roles such as technicians that maintain these automated trains. As accidents go down, the unions assume less risk for their members as well, as the unions often represent their members in cases of negligence. In much the same way that unions are opposing automated trains, insurers are opposing driverless cars. While accident rates will theoretically be reduced, they lose revenue as less people are buying their services (assuming the insurance risk with driverless cars shifts to the manufacturing corporations instead of the individual). While it is in their best interest to increase revenue, shouldn’t they also be motivated to decrease accident rates?

  2. Great post! I agree with Greg’s comment above that one of the biggest barriers of NJ Transit moving to driverless trains is the people. The Pennsylvania trains system (SEPTA) has only very recently implemented an RFID card ticketing system that has been the technology for many public transit systems for years. Prior to this, you had to purchase tickets through cash or credit at a sales counter through a clerk. Then this ticket was checked onboard by the conductor. This was a antiquated and terribly inefficient system, and despite the technology being available, SEPTA took a very long time to adopt it. The public has frequently blamed the union for pushing back on these changes. Similar to NJ Transit, SEPTA has experience a few strikes that severely disturbed travel for many. I’m sure that there is likely many states with similar issues. In my opinion, we are not close to a future where states will adopt driverless public transportation systems. However, I would like to see privately owned train systems such as Amtrak adopt such a change first.

  3. That is a great question and great idea. We should go automated, its a no-brainer. I’m pretty sure in non-right-to-work states its nearly impossible to fire employees represented by a union without facing major legal ramifications. Could it be that these labor inefficiencies actually slow innovation like self-driving trains?

  4. Interesting article, TUB.

    I think your article poses a difficult question and conundrum. On even a larger scope, I wonder how more predominantly unskilled laborers will feel and react to their jobs being replaced by machines. The benefits are very obvious with (eventually) lower costs after a large upfront capital expenditure as well as more predictability and safety in may industries. I think another added benefit, especially in this example, is the fact that with automated trains, they can run 24/7 without the necessity of overtime, so utilization can really increase. I would love to be able to take a train in the middle of the night when I would be sleeping anyway and wake up at my destination. I also agree with Greg that the loss of jobs is not as of too much concern since new jobs will naturally arise with the new technology [1].

    [1] Jones, Michael. “Yes, Robots will Steal our Jobs. And that’s Fine.” February 2016. Washington Post. https://www.washingtonpost.com/posteverything/wp/2016/02/17/yes-the-robots-will-steal-our-jobs-and-thats-fine/?utm_term=.8b302cf1002a, accessed November 2016.

  5. Really interesting topic! As a past frequent user of NJ Transit, this post really resonated with me, especially in the context of the latest accident in Hoboken about a month ago (that’s actually my home line). I recall reading an article that claimed that investments in more modern safety technology have the potential to have prevented that crash too.

    I’m not sure the funding/capital investment question is as simple as presented in the post. I actually see the availability of financing as the primary obstacle to implementation, not the union. The way that the US funds its public transportation is very localized (and unfortunately, therefore inefficient in my humble opinion). If more federal funds were available to fund these improvements (which ultimately,as you note, will improve safety and save lives!) I think we would see much faster adoption and implementation of automated trains.

Leave a comment