While still working on train automation, Metro makes moves toward a new signaling system

It’s been twelve years since Metro trains were primarily computer-operated, not driven manually by human operators. The fatal Fort Totten crash in 2009 led Metro to disable the system, only later to find out that it wasn’t the cause (a different part of the agency’s signal system was). By then, agency priority changes and time elapsed meant the system stayed off. Four General Managers later, the agency appears to have viable paths towards not just reenabling Automatic Train Operation (ATO), but potentially a giant signal system replacement, assuming groups inside the agency play nicely with each other.

Automatic Train Operation wasn’t intended to be off this long

The Metro system was designed with automation in mind and opened in 1976 with ATO as the primary mode of train operation during morning and evening rush hour periods.

June 22, 2009 changed all that. Inbound Red Line Train 112 crashed into the back of another train holding outside the Fort Totten station, killing eight passengers and the operator of the striking train. Within a day of the incident, without waiting for the lengthy National Transportation Safety Board investigation to wrap up, Metro took several reactionary steps that included turning off the ATO system, instructing all train operators to switch to manual operations.

Disabling ATO was publicly described as a temporary measure, a suspension that could be undone at some point in the future. “The ride will not be as smooth,” cautioned then-Metro General Manager John Catoe, “but this precaution is necessary until the investigation and our review of the system operation provide us more information.”

The agency stayed mum about when ATO might return, even after the NTSB determined it wasn’t at fault, and the system has remained mostly off. An effort to restore ATO began in 2014 and resulted in the system briefly being reenabled on the Red Line in 2015 for a few months. It was soon disabled again.

Metro had “more pressing concerns” to attend to, it said in 2017. The death of a rider, back-to-back derailments in 2015 and 2016, the resignation of the agency’s Chief Safety Officer, and a new General Manager led the agency to prioritize the safety of the tracks its trains ran on. The SafeTrack program was born soon thereafter to replace miles of track, cables, and thousands of rail ties and fasteners.

Now, Metro is considering two paths that could finally bring ATO back in some form.

Path number one: Work to restore ATO on the existing system

Metro actually does have plans to restore ATO. The system provides a smoother ride for passengers, and can reduce the workload of the rail controllers who govern the movement of trains. Restoring ATO while maintaining the existing equipment provides benefits while limiting the scope and cost of work that Metro would need to perform.

The to-do list standing between Metro and a return to ATO isn’t short.

ATO has been off long enough system-wide that the amount of time elapsed presents its own problems. Many train operators that have used ATO outside of training are no longer operating trains; the Automatic Train Control (ATC) departments that engineer and maintain the signaling system are struggling to stay fully staffed; new safety procedures protecting workers on the tracks conflict with how ATO is managed and rule changes are also needed. And, most recently, a Washington Metrorail Safety Commission (WMSC) oversight safety audit reported that groups that should be working together on returning ATO to normal operation aren’t collaborating.

A 2018 report by engineering firm Mott Macdonald laid out a number of steps Metro would need to implement before returning to ATO, including new training, procedure and rule modifications, and train and wayside equipment verification.

Metro appears to have followed those recommendations. A presentation prepared for the agency’s upper management in 2020 laid out a 17-month process that could culminate in the restoration of all-day ATO as early as next year, although the WMSC audit appears to have put a damper on that for now.

The WMSC found several issues with the way Metro was going about trying to restore ATO, including that the agency wasn’t following its own safety certification practice before turning the system back on. Additionally, groups that need to work with each other — the ATC Engineering group in the engineering department, the ATC Maintenance group in the rail maintenance department, and the Signal System Renewal Program (SSRP) group, in the Capital Delivery program — “do not appear to be on the same page or even aware of some actions being taken by other departments.”

Metro spokesperson Sherri Ly said the system is working on restoring Automatic Train Operation and Automatic Door Operation in coordination with the WMSC, but does not have a timeline. “Our current priority is on projects essential to safety and reliability,” she said.

The WMSC said in an audit it published in May, 2021 that the safety certification process needed before ATO could be restored “has not begun.” The SSRP group running ATO’s restoration also could not provide “documentation that hazards have been identified and mitigated” to certify that restoring ATO to service would be safe.

Path number two: Maintain the existing system, regardless of ATO restoration

While Metro attempts to restore ATO, the reality of age is presenting the agency with another problem: its existing train control equipment needs to be upgraded, regardless of future automation intent. Maintaining the existing “old” train control system is needed so the agency can continue to safely run trains on its tracks.

The equipment the agency needs to replace is located across dozens of Train Control Rooms situated along Metrorail tracks. Each room sends and receives all the signals that keep trains spaced and on the move in that area, and they all link back to the Rail Operations Control Center (ROCC) to provide rail controllers with a birds-eye view of the rail system.

The agency signaled (no pun intended) to companies that sell train control equipment it was preparing to make a large order to replace some of the equipment that’s now aging out. That existing train control equipment, which is between 21 and 44 years old, is not only wearing out, but has been superseded by newer technology.

Metro failed to anticipate and plan for the need to eventually replace its ATC train control room equipment, according to the WMSC, which is part of the reason the agency now needs to upgrade so much. Because the agency lacks a “standardized process to prioritize and advance ATC capital projects,” Metro didn’t develop plans for how and when to perform replacement until they were facing the imminent need to do so. Information available to the ATC Engineering group that told them about failures, delays, and expected equipment lifespans wasn’t always acted upon.

According to Metro budget documents, the first phase would cost around $83 million. Additional phases of the project, replacing the rest of the equipment, would cost around $180 million.

Metro expects the first phase of the Train Control Room refresh project to take three years; the second, larger phase would presumably take longer.

The agency can’t — or at least doesn’t want to — delay this project while figuring out its signaling future and risk not being able to keep the system maintained. Without maintaining the train control system, the WMSC warned, “Metrorail risks a catastrophic accident as occurred in 2009 and 2015.”

The actual contracts to replace the equipment have not yet been awarded.

Path number three: Start over with a new signaling system

The agency has for a few years now been presented with a decision: do they spend money to return the existing system to automatic train operation, do they dive in and upgrade to the latest widely-accepted signaling system, or some combination of the two?

A 2020 study by Metro staff and an engineering partner found that rather than continuing to maintain the existing system, Metro’s best option could be to replace the signaling system with Communications-Based Train Control (CBTC), a completely new system that relies on wireless radio transmitters to keep track of trains and send and receive information.

Another alternative would be to upgrade the existing system from analog relay-based circuits to digital track circuits. Both options would provide benefits to Metro, the report said. The two alternate systems would provide more flexibility to the control center, be more reliable, and would include additional redundancy that Metro’s current system doesn’t have.

The CBTC upgrade option out-scored the first two options on 9 of 11 assessed “key performance indicators,” falling behind only on “impact of migration” and “modifiability.” The report found CBTC would improve safety and security, system capacity, and operational flexibility (like being able to use automation in single-tracking areas, which Metro currently can’t).

CBTC would also set Metro up for additional future train automation, even allowing for trains to run without operators one day, the report concludes.

Would be the ideal opportunity for WMATA to convert its trains to driverless operations & install platform barriers, as Paris has done on existing lines, which would allow more frequent & faster service + reduce accidents. https://t.co/8M5QpbFnsH

— Yonah Freemark (@yfreemark) August 10, 2021

Still, CBTC might not make that much of a difference in capacity. An analysis from Metro staff in 2017 estimated that CBTC might be able to provide a “modest 1-2 trains per hour” throughput increase from the current limit of 26 trains per hour, but couldn’t go much further due to other constraints.

Metro seems interested in implementing CBTC. To learn more about the option, in August Metro issued a Request for Information, an open invitation for vendors to send the agency information about their signaling system products.

CBCT could also have an edge over upgrading the existing track circuits to digital because it would take less time — just 10 years — and wouldn’t require single-tracking. Upgrading from an analog to digital ATC system would be more complex, requiring signaling equipment upgrades at every train control room throughout the rail system, and would purportedly take 16 years to upgrade the entire system.

“No determination has been made with regard to CBTC,” said Ly, Metro’s spokesperson, adding: “Any solution will involve large-scale upgrades that require long-term capital construction impacting service.”

Metro documents confirm that details regarding the implementation of CBTC at Metro are still in development, but include language that indicates the agency is setting up working groups to determine specifics.

While Metro embarks on figuring out what it might take to upgrade the system to CBTC, paying for it is another problem. Metro’s initial estimate pegs a system-wide CBTC upgrade at “over $2 billion,” and the agency only has $13 million programmed in its 6-year Capital Improvement Program that runs from 2021 through 2027.