Engineering The Subway That Built a City

Films love to show New Yorkers chasing cabs down busy streets, but the real secret to getting anywhere fast has always been underground—ever since the Interborough Rapid Transit (IRT) made its debut.

On October 27, 1904, crowds gathered at the (now decommissioned) City Hall station to board the very first trains of the IRT, or as we know it today—the subway. The inaugural 9.1-mile route extended from City Hall to 145th Street Harlem, promising what seemed impossible at the time: an uptown ride in just 15 minutes.

“Imagine standing at City Hall that day. People weren’t sure if they’d be safe underground, but they were curious enough to try,” said Jodi Shapiro, curator at the New York Transit Museum in Brooklyn. “That first ride wasn’t just about getting from point A to point B—it was about believing that technology could reshape daily life.”

More than four years of construction had passed since the groundbreaking ceremony in 1900, and engineers had pulled off an extraordinary feat: carving tunnels below the busiest streets in America without bringing daily life to a halt. The scale and ingenuity of the project would later earn the Interborough Rapid Transit System its designation as an ASME Historic Mechanical Engineering Landmark.

 

A city at standstill

As with so many engineering marvels, New York’s subway system was born out of necessity. By the turn of the 20th century, the city was choking on traffic: Horses, trolleys, and elevated trains could no longer keep up with a population that had nearly doubled in a single generation. The catch-22 was finding a solution that could relieve the chaos without adding to it.

Other cities had already turned underground, Shapiro shared. “New York was actually late to the game. London and Boston had shown it could be done, but New York demanded something bigger and faster,” she said, referring to London’s Underground and the Tremont Street Subway. 

To meet the crisis, the city commissioned William Barclay Parsons, chief engineer of the Rapid Transit Commission, and contractor John B. McDonald to build a system unlike anything attempted in Manhattan before.
 

Excavating Manhattan

Engineers were faced with a maze of urban constraints. Above, there were building foundations, elevated railways, and heavy surface traffic. Below, the sewers, water and gas mains, power conduits, telegraph lines, and even vaults extending from private basements under the street. 

“You’re digging these tunnels underground to move people. You do have to move some utilities around,” Shapiro explained. “But it was also an opportunity for New York to make improvements. Part of what was built into the subway construction contracts was to remake the sewage system in lower Manhattan.”

Every section of the tunnel was an exercise in conflict resolution. Chief Engineer Parsons recommended shallow excavation wherever possible, drawing on the Boston subway as precedent. The so-called “arcade” design used steel I-beams with concrete arches between them, waterproofed and backfilled before traffic returned overhead. More than half of the first IRT line was built in that manner.

The contractor, McDonald, is quoted on multiple occasions emphasizing the monumental task of digging a great trench through the heart of the most crowded city of the world without disturbing the traffic and every day business of the people: “Imagine a surgeon who has to direct a probe from the neck of a patient all the way to each heel... there you have a fair likeness of what had to be done in digging the rapid transit tunnel.”

But Manhattan’s geology and topography forced variety. Cut-and-cover trenches worked in Lower Manhattan’s softer soils, while uptown, solid schist allowed traditional rock tunnels. Under the Harlem River, engineers advanced shield-driven cast-iron tubes under compressed air, an advanced method for its time. In topographical depressions, steel viaducts carried the line above ground, while on Lenox Avenue, engineers embedded steel rods in concrete instead of heavy beams.

“New York needed all of those methods, depending on where the line was going to run,” Shapiro explained. “Manhattan’s rock formations are unusual—twisted and folded almost like a pile of laundry thrown on the floor, as one geologist put it. 

“You could excavate a section that looked fine, and then just a few feet later the rock was shearing off in another direction,” she continued. “Around 42nd Street, near Grand Central, the conditions were especially difficult, and engineers described it almost as a cursed section.”

Expansion of the system

The success of the first IRT line was immediate. Ridership soared into the hundreds of thousands, and expansion followed at a staggering pace. “Even while that first route was being constructed, plans for other routes were being brewed up,” Shapiro said. 

By 1908, new lines were already carrying passengers to the Bronx and Brooklyn. Elevated viaducts and deep rock tunnels connected distant neighborhoods to the original line, bringing the city together like never before and redefining where New Yorkers could live, work, and imagine their futures.

The Dual Contracts era in the 1910s and 1920s represented the most ambitious building phase, extending service far deeper into the outer boroughs. “That was the biggest expansion of subway service,” Shapiro said. “The next real transformation happened with the Chrystie Street Connection in the 1960s. It changed how the system worked—more of an organizational shift than a physical one, but still a game changer.” The IRT eventually merged with other operators, creating the framework for today’s vast city-owned MTA system.

Building more tunnels was only part of the challenge—engineers had to scale the entire ecosystem of subway technology to keep pace with a city that refused to stop growing. And the work continues: today, the MTA is advancing plans for the Interborough Express, a new rail link between Brooklyn and Queens, echoing the same spirit of expansion that defined the subway’s earliest decades.
 

Subway tech

The subway was never just about tunnels and stations. The mandate of the contract under which the first subway line was built stated that the contractor was required to provide “a complete equipment for the railroad, including not only cars, but also engines, electric wires, conduits, power houses, and lighting, signaling, and ventilation apparatus.”

From its opening day, the IRT was a fully integrated system of rolling stock, power delivery, signals, and fare control—each engineered to work together at a scale no city had attempted before. Many of these technologies were groundbreaking at the time and, remarkably, some of their basic principles still govern subway operations today.

Early subway car design

The first IRT cars were designed to be safer, sturdier, and more comfortable than anything New Yorkers had seen on the elevated lines. Built with all-steel underframes and anti-telescoping bulkheads, they were engineered to resist crushing in collisions—a dramatic improvement over wooden cars.

Enclosed vestibules with sliding doors replaced the open, gated ends of earlier trains, protecting passengers and speeding up boarding. At 51 feet in length with seating for 52, these cars were larger and more robust than their elevated predecessors, setting a new physical standard for rapid transit rolling stock.

Inside, passengers found bright electric lighting, improved ventilation through ceiling vents and operable windows, and electric fans that circulated air through the crowded cars. The electrical systems were standardized so that any car could serve as either a motor or a trailer, giving engineers flexibility in train assembly. 

Performance was equally impressive: the fleet’s motors delivered nearly 2,000 horsepower for express service, enabling trains to reach 55 miles per hour. “It’s called the Interborough Rapid Transit. To be rapid transit, it has to go over fifty-five miles an hour,” Shapiro clarified. “And it did from the very start.”

The IRT’s cars were never static. Each generation reflected advances in materials, passenger expectations, and system demands.

1910s to 1930s: Standardization and safety. The IRT introduced “Lo-V” (low-voltage) cars, which isolated the high-voltage third-rail power from operators through relays, a major safety innovation for motormen and maintainers. Cars grew longer with wider doors to speed boarding as ridership exploded.

1940s to 1950s: Stainless steel. Pioneer trains like the R10 and R11 experimented with stainless steel, reducing weight and corrosion. These set the pattern for mid-century rolling stock across U.S. transit.

1960s to 1980s: Passenger comfort. Ceiling fans gave way to forced-air ventilation and, eventually, air conditioning. Interiors used fiberglass seats and fluorescent lighting. The shift was as much about efficiency and cleaning as it was about comfort.

1990s to 2000s: Digital control. Cars were equipped with electronic diagnostics, digital signage, and automated announcements. Modular truck and braking systems simplified maintenance.

Modern day: Open, connected, automated. The newest models, including the R142, R160, and the pilot R211 cars, feature open gangways between cars, crashworthy cab designs, and communications-based train control (CBTC) interfaces. Onboard microprocessors monitor traction, braking, and doors in real time.



“Every generation of subway car looks a little different—the seats, the grab bars, the interior layout,” Shapiro said. “The tunnels and stations don’t change much, but the cars themselves evolve constantly.”
 

Third rail and power delivery

Equally critical to the IRT’s success was its choice of power. Engineers rejected overhead wires as impractical for underground operation and instead adopted a 600-volt direct-current third rail system—one of the first large-scale applications in the United States. A steel conductor rail was mounted alongside the running rails, covered by wooden protection boards, and energized by substations spaced along the line.

Each car was equipped with sliding contact shoes that pressed against the third rail to draw current, while the running rails returned the circuit to ground. This design eliminated smoke and soot from steam engines, making underground service feasible, and it set the standard for electric rapid transit worldwide.

The system’s reliability depended on a network of powerhouses and substations, where rotary converters transformed high-voltage alternating current from the grid into the low-voltage direct current required by the trains. These substations, located roughly every two miles, became as essential to the subway as the tracks themselves.
 

Signaling and interlocking

The IRT was engineered with one of the most sophisticated signaling systems of its time. Designed by the Union Switch & Signal Company, it combined track circuits, electro-pneumatic interlocking, and automatic train stops into a unified logic network that allowed local trains to run every 90 seconds and express trains every 120 seconds. 

The genius of the system was its layered safety: electrical detection, overlap redundancy, mechanical train stops, and interlocking logic all reinforced one another. This fail-safe design gave New Yorkers confidence in the world’s first high-density rapid transit, and its principles still underpin subway signaling more than a century later.



Track circuits and block signals
The subway line was divided into electrically isolated “blocks,” typically 600 to 1,000 feet long. Current flowed through the rails in each block; a train’s axles shorted the circuit, which dropped the protecting signal to “stop.” This was a fail-safe design: loss of current, whether from a train, a broken rail, or faulty wiring, always resulted in a red signal.

Headway equations
Engineers calculated that with blocks 600 feet long, and trains averaging 30 mph (44 feet/sec), safe braking distance required 2 to 3 blocks of separation. This math fixed the theoretical maximum headway of 90 seconds for locals and 120 for expresses, a capacity target that the IRT delivered in practice.

Overlap protection
To further prevent rear-end collisions, the IRT adopted the British practice of overlapping blocks. A train in one block also held the signal behind it at red until it cleared part of the next block. This redundancy meant a following train could never legally approach closer than several hundred feet, even at high speeds.

Automatic train stops
Every wayside signal was coupled with a mechanical trip arm. If a motorman ran a red, the raised arm engaged a valve on the train’s brake pipe, instantly venting compressed air and applying the brakes. This integration of wayside signals with rolling stock braking was a pioneering fail-safe mechanism.

Electro-pneumatic interlocking
At junctions and terminals, towers controlled switch points with electro-pneumatic machines: an electrical command from the lever frame energized a magnet, which opened a valve, which directed compressed air to move the switch. Each lever was wired into interlocking logic: no conflicting routes could ever be set. As curator Jodi Shapiro put it: “One of the cool things about an interlocking is it’s just yes or no.”

Power interference solutions
Because the rails carried the return current from the 600V DC third rail, ordinary DC track circuits would have been swamped. Engineers instead adopted alternating current track circuits, separated by tuned relays, so train detection remained reliable.

Longevity of electromechanical relays
Many original signal relays and interlockings, over a century old, still function today. “We’ve got relays in our system that are one hundred and twenty years old, and they’re still operating and they’re fine,” Shapiro noted.

Transition to CBTC
Today, Communications-Based Train Control (CBTC) is gradually replacing fixed-block signals with a moving-block system. Instead of rigid blocks, trains report their precise location via radio, allowing shorter headways and higher capacity. The L and 7 subway lines are now fully CBTC, while others are in progress.

 

Turnstiles and fare control

Just as tunnels, tracks, and signals had to be engineered for heavy use, so too did the act of collecting fares. From the get-go, the IRT has relied on machines designed for reliability and endurance under millions of daily cycles. And yes—for as long as there have been fares, there have been efforts to avoid paying them.  

Here’s a look at how fare control has evolved over the years.

Ticket choppers (1904): At opening, passengers purchased paper tickets (for a nickel) which were validated by a mechanical “chopper.” The device sliced a corner off each ticket, preventing reuse. Simple but effective, it kept the flow of riders moving while reducing fraud.

Turnstiles (1920s): By the 1920s, the chopper gave way to the iconic rotating turnstile. Each passage rotated a locked arm, releasing only when a nickel was deposited. These mechanical devices were built for extraordinary durability, cycling hundreds of thousands of times each week without failure. Turnstiles soon became synonymous with the subway itself, an engineering solution scaled to human movement.

Tokens (1953 to 2003): As fares rose beyond a nickel, paper tickets became impractical. Engineers developed robust coin-handling mechanisms to accept specially minted brass tokens. The machines were simple, rugged, and remarkably secure.

MetroCard (1993–2024): In the digital age, magnetic-swipe MetroCard readers replaced tokens. Though faster and more flexible, they required regular calibration and maintenance.

OMNY (2019–present): Today, contactless smart cards and phones can tap into OMNY readers, marking the shift to fully electronic fare collection.
 

A Lasting Legacy

Like all great engineering systems, the subway was built to endure—so seamlessly that its longevity can be easy to take for granted by today’s commuters. “New York wouldn’t be the city it is without public transportation,” Shapiro said. “The subway opened up the flesh of New York and inserted this new vein.” Many of those original tunnels and stations remain in daily use more than 120 years later. 

As Parsons predicted in 1900, “by the time the railway is completed, areas that are now given over to rocks and goats will be covered with houses, and... a special traffic of its own." The IRT was not only a transit solution but the framework for New York’s expansion. Today, North America’s largest transit system remains an engineering marvel—carrying millions across the city it once helped build.

Sarah Alburakeh is strategic content editor.