The Boring Company builds its own tunnel boring machines, runs its own tunnels and fills them with someone else’s cars. Its supply chain is unusually short on the machine side and unusually dependent on the vehicle side — and in 2026 the real constraint is neither. It is permits. Here is the full chain, from the in-house Prufrock TBM to the Tesla fleet that makes the whole system move.
👉 The counterintuitive core of this story: The Boring Company solved the hard engineering problem — building a fast, reusable, electric TBM — and got stuck on the boring one. Paperwork, not geology, is what gates the 68-mile network.
- The machine: Prufrock is designed and built in-house — The Boring Company is a TBM manufacturer, not a TBM buyer, which is rare in tunneling.
- The tunnel: lined with precast concrete segments installed continuously as the machine advances, not in stop-start cycles.
- The vehicle: the entire passenger experience runs on Tesla cars — a total dependency on a sister company.
- The progress: Prufrock-2 completed a record 2.28-mile drive in March 2026; roughly 11 miles bored, about 3.5 miles operational.
- The real bottleneck: not steel or concrete — permitting. Vegas Loop needs around 600 permits and had just over 100 in early 2026.
Who Builds the Machines? The Boring Company Does
In conventional tunneling, the boring machine is the biggest single procurement item. Firms buy custom TBMs from a handful of specialist manufacturers — Germany’s Herrenknecht is the dominant name, supplying machines to as many as 100 projects a year at its peak (background on the global TBM market).
The Boring Company broke that model on purpose. Rather than order a machine tailored to one dig, it designs and iterates its own Prufrock line and reuses each machine across tunnels. That single decision is why its supply chain looks nothing like a traditional contractor’s: the most expensive, longest-lead component is internal.
What makes Prufrock different
Prufrock’s design targets the three things that make traditional TBMs slow. It installs precast segments while advancing instead of stopping every five feet, it "porpoises" — launching into the ground from a truck within 24 hours rather than needing an excavated pit and cranes — and it is engineered for Zero-People-In-Tunnel operation controlled from a central center (The Boring Company’s Prufrock page).
The company also swapped rail-based spoil removal for rubber-wheeled segment trucks and runs the machines on electricity rather than diesel, simplifying ventilation (NextBigFuture on the Prufrock design changes). The stated goal is one mile per week — an order of magnitude faster than industry norms.
💡 Reality check on the headline number: independent tunneling analysts note the "1 mile per week" figure has not been publicly demonstrated across varied geology. The March 2026 record of 2.28 miles was a single drive, not a weekly rate. Treat the target as a design goal, not an achieved benchmark.
The March 2026 Milestone — and What It Reveals
On March 9, 2026, Prufrock-2 emerged after boring a 2.28-mile tunnel near Westgate — the longest single Vegas Loop drive to date, beating the company’s previous 2.26-mile record. The bore moved roughly 68,000 cubic yards of dirt using about 4.8 miles of continuous conveyor belt driven by six motors totalling 825 horsepower (Teslarati on the Prufrock-2 record).
The conveyor detail is the tell for supply-chain watchers. Nearly five miles of belt per tunnel is a serious consumable, and continuous spoil handling is one of the least-glamorous, most-critical parts of the chain. Total tunnel construction stood at roughly 11 miles by mid-2026, of which about 3.5 miles were operational, with the rest awaiting station readiness and regulatory approval (Vegas Loop progress tracker).
✔ Note the gap: 11 miles bored, 3.5 operational. The tunnels are ahead of the stations and the permits. That imbalance tells you exactly where the chain is constrained — not in the ground, but at the surface and in the paperwork.
The Four-Part Supply Chain
The Boring Company’s inputs sort cleanly into four categories, each with a very different sourcing profile and risk.
| Segment | What it covers | Sourcing profile |
|---|---|---|
| The machine | Prufrock TBM: cutterhead, drive motors, hydraulics, control systems | In-house design and build; components from industrial suppliers |
| The tunnel | Precast concrete lining segments, conveyor belt, grout | External — concrete and belt are the main recurring consumables |
| The system | Stations, ventilation, lighting, power, comms, fire safety | External construction and MEP contractors |
| The vehicle | Tesla cars, charging, eventual autonomous operation | Tesla — single-source by design |
The precast concrete question
Every mile of tunnel needs a full ring-set of precast concrete segments. This is the largest recurring material input in the entire operation — far more tonnage than the machine itself over a project’s life. Segment production is either done in a dedicated on- or near-site casting yard or bought from regional precast suppliers; for a 68-mile network, segment logistics is a bigger sustained challenge than boring speed.
💡 Underappreciated cost driver: at roughly $10 million per mile for tunneling alone (per Boring Company’s own figure), and with 68 miles approved, precast concrete and its transport represent one of the largest line items in the whole program — and one of the few the company cannot iterate its way out of.
The Tesla Dependency: A Feature and a Risk
The Boring Company does not make vehicles. Every passenger ride in the Vegas Loop happens in a Tesla, and the long-term vision — up to 1,200 vehicles across a 68-mile network — assumes a Tesla fleet operating autonomously in a closed environment (coverage of the Vegas Loop expansion plan).
This is vertical integration across sister companies rather than within one firm. It is efficient — the Loop becomes a live proving ground for Tesla autonomy, and Tesla supplies a fleet at cost — but it is also a single point of dependency. The Loop’s throughput ceiling, vehicle cost and autonomy timeline are all set by decisions made at Tesla, not by The Boring Company.
| Vegas Loop capacity metric | Figure (2025–2026) |
|---|---|
| Total passengers to date | Over 4 million since 2021 |
| Peak throughput demonstrated | Around 32,000 passengers per day; 4,500–6,600 per hour |
| Current operational tunnel | Roughly 3.5 miles, 11 stations |
| Approved full build | 68 miles of tunnel, 104 stations |
| Full-build target capacity | About 90,000 passengers per hour |
| Full-fleet target | Up to 1,200 Tesla vehicles |
✔ The dependency cuts both ways. If Tesla solves unsupervised autonomy, the Loop’s economics transform overnight — no drivers, dramatically higher throughput. If Tesla’s autonomy slips, the Loop stays driver-limited regardless of how many tunnels get bored. The vehicle chain, not the tunnel chain, controls the upside.
The Real Bottleneck Is Permits, Not Parts
Here is what most coverage misses. The constraint on the Vegas Loop is not TBM availability, concrete supply or labour. It is permitting. The project requires roughly 600 permits across its 68 miles, and as of early 2026 the company held only slightly more than 100 (NVBEX on the permitting constraint).
Each tunnel segment, each station, each surface connection needs its own approvals across multiple agencies. The airport connection alone required an Automated Vehicle Identification permit from the Clark County Department of Aviation, subject to documentation review, insurance verification and inter-agency coordination (Review-Journal on the airport permitting). The full network is tentatively scheduled for completion between 2028 and 2029.
⚠ For anyone modelling Boring Company timelines: bore rate is the wrong metric. A machine that can theoretically dig a mile a week is irrelevant if permits arrive at a hundred every few years. The rate-limiting step is administrative, and it does not respond to better engineering.
Prufrock vs. Traditional TBMs: The Supply-Chain Contrast
| Dimension | The Boring Company (Prufrock) | Traditional (e.g. Herrenknecht) |
|---|---|---|
| Machine sourcing | Designed and built in-house, reused | Custom-built per project by specialist maker |
| Lining install | Continuous, simultaneous with boring | Stop-start cycle every ~5 feet |
| Launch method | Porpoising from a truck in 24h | Excavated launch pit with cranes |
| Spoil removal | Rubber-wheeled trucks, electric | Rail locomotives, often diesel |
| Crew in tunnel | Zero-People-In-Tunnel design | Crewed operation |
| Diameter focus | Small (car-width) tunnels | Small to very large diameters |
| Best suited to | Repeated tunnels, consistent geology | Bespoke, high-consequence public works |
The trade is real and worth stating fairly. Prufrock optimizes for repeatability and speed in known, consistent geology — exactly the Las Vegas use case. Herrenknecht machines are tailor-made for the hardest, most variable conditions in large public works, where a proven, slower approach is preferred (Prufrock vs Herrenknecht comparison). One is a product company; the other is a bespoke engineering house.
💡 The strategic insight: The Boring Company is trying to turn tunneling from a project business into a product business — standardized machines, standardized tunnels, standardized cost. That is the same playbook Tesla and SpaceX ran. Whether it works depends on whether geology stays consistent enough to standardize, which it does in Vegas and may not elsewhere.
What the Expansion Pipeline Implies for the Chain
Beyond Las Vegas, the company has floated one-mile demonstration tunnels in Henderson and Summerlin, and President Steve Davis raised the idea of a roughly 440-mile hyperloop link between Las Vegas and Reno at an estimated $100 million (NVBEX on expansion plans). Prufrock-3 completed a tunnel to Giga Texas, and Prufrock-4 has entered testing, signalling a growing machine fleet (Teslarati on the Prufrock fleet).
Each new geography changes the supply chain in one specific way: geology. Consistent soft soil in Las Vegas is what makes standardized precast segments and a fixed machine design work. Different ground means different linings, different cutterheads and a re-opened supplier question the Vegas model had closed.
Checklist: How to Evaluate a Boring Company Claim
- Separate bore rate from delivery rate. A record single drive is not a sustained weekly rate, and neither is the true constraint.
- Count permits, not miles. Progress is gated by approvals; ask how many of the ~600 are in hand.
- Check the operational-vs-bored gap. Tunnels bored but not open signal a surface or regulatory bottleneck, not an engineering one.
- Follow the Tesla autonomy timeline. Loop throughput economics depend on it more than on tunnel count.
- Watch geology on any new site. Consistent ground is what makes the standardized supply chain work; new terrain reopens it.
FAQ
Who makes The Boring Company’s tunnel boring machines?
The Boring Company designs and builds them in-house. The Prufrock line is proprietary and reused across tunnels, unlike traditional contractors that buy custom machines from makers such as Herrenknecht.
What is the Vegas Loop built from?
Tunnels lined with precast concrete segments, served by surface and subsurface stations, and operated with Tesla vehicles. Precast concrete and conveyor belt are the main recurring material inputs.
Does The Boring Company depend on Tesla?
Completely, for vehicles. Every passenger ride uses a Tesla, and the full-build vision assumes up to 1,200 Teslas operating autonomously. The Loop’s throughput ceiling is tied to Tesla’s autonomy progress.
What is actually slowing the Vegas Loop down?
Permits. The 68-mile network needs roughly 600 permits and had only about 100 in early 2026. Tunnels have been bored faster than stations and approvals can catch up.
How fast can Prufrock dig?
Its design goal is one mile per week. The demonstrated record is a single 2.28-mile drive completed in March 2026; a sustained weekly mile has not been publicly shown across varied geology.
When will the full Vegas Loop be finished?
Tentatively between 2028 and 2029, contingent on permitting and station construction rather than on tunneling speed.
The Bottom Line
The Boring Company’s supply chain is a barbell: radically vertically integrated on the machine, radically dependent on a sister company for the vehicle, with fairly conventional concrete and construction in between.
It solved the part everyone assumed was hard — building a fast, reusable, electric, crewless TBM — and ran into the part nobody puts on a keynote slide: 600 permits, station construction and inter-agency coordination. The machine is ahead of the paperwork by design, and the paperwork is what ships tunnels.
Judge progress in 2026 on three numbers only: permits secured against the 600 needed, operational miles against miles bored, and the Tesla autonomy timeline that sets the Loop’s real capacity. Everything else is dirt already moved. Track how this connects to the wider Musk ecosystem in our Tesla supply chain breakdown.
💡 We track the whole Musk hardware ecosystem — Tesla, SpaceX, Neuralink, The Boring Company — with confirmed and reported figures kept in separate columns. Join the community to get each supply-chain breakdown as the news lands.
Sources: The Boring Company (Prufrock and Vegas Loop pages), Teslarati, Tesla North, Review-Journal, NVBEX, Tesorb, NextBigFuture, Resource Erectors, Vice. Figures reported by third parties are noted as such. Last fact-check: August 5, 2026.
NEVER MISS AN OPTIMUS UPDATE
We track every Tesla Optimus development — specs, deployment milestones, pricing and competitive moves — updated as news breaks.