Boring Company’s Vegas Loop Is Quietly Building America’s Biggest Transit System

With tunneling under $10 million a mile, point-to-point Teslas, and airport lines nearly live, the network already moves stadium-scale crowds and is racing toward top-tier U.S. ridership.

The Boring Company’s Vegas Loop is not a demo under a convention hall. It is already one of the larger public transportation systems in the country, and on a busy convention day a single stretch moves about 40,000 people - more than most American light-rail systems. Two new tunnels, among the longest transport tunnels built in the U.S. in the last 25 years, are set to multiply ridership, and the full build-out is aimed at a top-three ranking behind only the densest legacy systems. The edge is simple: dig cheap, keep stations tiny, and let individual cars run express so every new stop does not slow everyone else down.

Key Takeaways

• Busy convention days already push about 40,000 riders through a small slice of the network - more than the majority of U.S. light-rail systems.
• American subway construction often runs $1–5 billion per mile; Loop tunneling is under $10 million a mile, with a path toward roughly $4–5 million as machines hit continuous mining and Z-PIT goals.
• Fountainbleau to Caesars can run about 15 minutes and $26 on surface rideshare versus roughly 2–3 minutes and about $4–$4.50 per person on Loop.
• Median wait system-wide is zero seconds; mean wait sits near 30 seconds, with more than half of riders not waiting at all.
• Opening the airport lines (targeted around August) flips the ride from mostly surface to about 90% tunnel and expands stations from roughly 11 to 19 overnight.
• Seven boring machines run today, with 11 expected by year-end; one foot of tunnel equals about one dump truck of dirt - a mile is 5,280.
• Stations fit in a few parking bays (often 3–4 spots; the biggest around 20), with the hard rule that riders exit within 100 feet of a property’s main door.
• About 120 property owners have signed entitlements; roughly 600 permits cover the full Vegas program, built segment by segment rather than as one mega-permit.

---

Dirt, Not Hype, Is the Scale Metric

By yards of dirt moved and concrete poured, Vegas Loop ranks as the largest infrastructure project in America - on the order of 10x a stadium. Most people still do not know it exists. That gap is the story: massive underground capacity is already operating while surface life barely notices the dig.

I find that scale hard to hold in your head until you stack it against what we usually call “big.” A stadium rebuild makes the news. Miles of tunnel under a city, with no road closures, barely registers. That is the point of building in three dimensions instead of fighting over the last lane of asphalt.

Point-to-Point Beats Stop-After-Stop

The Strip is a bad fit for traditional transit. Airport to the top of the corridor as a bus or train with a stop at every major property becomes a 90-minute crawl. Cut stops and you lose flexibility. Keep them and you punish every rider who is not getting off.

Loop’s design answers that with individual vehicles - Teslas - and an express model. A given origin-destination pair stays about eight or nine minutes whether or not intermediate stations exist in between. Adding a station does not lengthen everyone else’s trip. That is why a station can be a resort portico or a taco-shop parking lot: the unit economics of a few parking spaces, not a multi-acre terminal.

Why Cars Underground Break the City Land Trap

Cities devote roughly 30–80% of land to cars - parking, lanes, the whole surface machine. People still want cars because cars give point-to-point freedom. Put that same vehicle underground and you keep the product people already choose while reclaiming surface for parks, pedways, and buildings that pay taxes instead of storing empty metal.

Bottlenecks get a blunt answer: dig another tube. Unlimited capacity is not a slogan when the cost and speed of tunneling stay low enough that another parallel segment is an engineering decision, not a decade-long political fight. Elon’s framing holds: we live in a 3D world and keep designing 2D transport.

The Real Secret Sauce Is Cost and Speed

Architecture and Tesla handoffs are the visible product. The moat is tunneling economics. Subway construction in America, ignore operations, often lands between $1 and $5 billion per mile. Loop is under $10 million a mile today and still moving.

Prufrock machines, largely built in Bastrop, Texas (early units partly outsourced; later ones near fully in-house), aim for continuous mining - push while you build the ring, instead of stop-start - and Z-PIT, zero people in the tunnel. Crew counts already fell from about a dozen on the first machine toward two people total behind the shield, with ops remote from Texas. Hit continuous mining plus remote control and unit cost starts to converge toward $4–5 million a mile. Rural highway lanes can run $2–5 million per lane; urban can hit $5–20 million. Below five, the question flips: why pave another surface lane when you can bury capacity and free the street?

I know how that sounds to anyone who has watched U.S. mega-projects. The counter is the ledger already in the ground, not a slide deck.

Stations That Fit Like Rideshare Spots

The 100-foot rule has not been broken yet: step out of the Tesla and you are within a hundred feet of the main front door - no Ryanair-style bus to the real destination. Large stations might hold about 10 parking bays; many run three or four. The convention center’s biggest footprint is on the order of 20 spots and still supports tens of thousands of daily boardings because load is spread across the map.

Core surface infrastructure to open a portal can sit near a thousand square feet; porpoising launches the machine from the surface with a small footprint (on the order of 14–15 by 70 feet) instead of a giant launch pit. Integrate into an existing garage the way you would reserve Uber stalls. Velocity, not acreage, is the capacity lever. A thousand people an hour at one station looks small until you have a hundred stations and almost nobody is going to the same place at once - the same reason Cybercab-scale vehicles fit average trips of about two people and the bulk of U.S. travel is two riders or fewer.

Annex stations extend the network without a full portal at every property: drive on public roads to the nearest tunnel mouth, drop in, and continue. A few of those already exist - including access patterns near Fountainbleau - so you are “in the system” the moment a Loop car can reach you.

Airport Lines, Permits, and the Clip-or-Bypass Game

Today an airport-oriented trip can still be mostly surface. When the new airport tunnels open, the pitch is about 90% tunnel and 10% surface, plus roughly eight stations along the corridor, taking the map from about 11 stations to 19 in one cut. Full network ambitions sit in the top tier nationally - second or third behind New York’s head start, in the conversation with D.C. and Chicago for scale once everything is built.

Vegas permitting is segmented - on the order of 600 permits for the whole program - versus Nashville’s single-shot program approval, where two machines are already working. Entitlements already cover about 120 property owners. When the machine approaches, owners choose clip or bypass. Clip means the tunnel sits under their land so a future station is a dig-down, not a fight for Strip right-of-way. Bypass means they stay off the map until they change their mind. Stations can open years later without slowing express traffic - fiber in the ground before you order service. Even a small coffee shop has already signed for a station with the same map weight as the big resorts.

Autonomy Makes More Traffic - Tunnels Are the Outlet

Cheaper autonomous miles do not empty roads. They fill them. More trips, more empty repositioning, more demand. Flying cars are a different engineering problem. Underground network capacity is the one already shipping.

FSD rides inside Loop tunnels are already part of the product story: controlled environment, no pedestrians, no weather, traffic management on an in-car tablet, green-button call to control. Median wait of zero seconds and mean near 30 seconds set a bar most surface transit never touches. A three-second glitch is treated as a routing bug to fix, not “normal.” That is the standard you get when the product is supposed to feel like a private car, not a schedule.

What Lives Under the Road Deck

Forty-five feet down you still get five bars of 5G. Under the roadbed: RFID and multiple navigation systems, comms, power, internet, drainage spanning 10-plus miles into wastewater treatment, and something like 10–20 thousand miles of fiber strands across the system. Cross passages and exits - hand-dug, harder than the main bore - sit where surface exits are not available. Wet standpipes with hose valves shaped for fire-glove access show the sales job is safety detail as much as speed.

Utilities belong in tunnels. Dig once, service without carving asphalt, flood a tunnel instead of a street. Loop has never closed a road for this build. Surface impact concentrates at the launch/muck site while the machine may be miles away. Continuous conveyors haul dirt; stripped Model 3 “liner trucks” (about 24,000-pound segments, short range, cameras everywhere) feed the face under remote control from Texas, with a local backup. One foot of progress is a dump truck of muck; the surface truck fleet is often the bottleneck - hence service tunnels for dirt so mining does not wait on Strip traffic.

Nashville Rock and the Delete Strategy

Vegas is mostly soft soil with caliche pockets. Nashville is limestone hard enough that much of the tunnel needs no liner - natural rock as the finish. That kills the push-off-the-segment model. The answer is gripping: millions of pounds of side force against the rock, mine, re-grip. Whole liner-handling subsystems disappear. Fewer sensors, faster digs, lower cost - the classic delete move. A soft-soil Prufrock can work in rock with a conversion kit, but a purpose-built main-beam machine wins when the geology allows you to throw away half the dance.

So the product thesis is clean. Keep the car people already love. Make tunneling cheap enough that capacity is an engineering dial. Scatter small stations so throughput multiplies without turning every trip into a milk run. Pair that with autonomy and the underground network becomes the only scalable answer to the traffic that cheaper miles will create. Vegas is already running the proof. The rest is machines, permits, and how many cities decide they want to be clipped before the next boom leaves them on the surface.