You can be forgiven if you don’t remember Arrivo, a shuttered Hyperloop-esque company founded by the former Hyperloop One executive Brogan BamBrogan. In addition to some wildly ambitious ideas, it had many slogans: “Engineered to get us there.” “The End of Traffic.” But the one that best encapsulated the erstwhile company’s disregard for the practical limits of space-time: “Don’t get there, be there.”
The company shut down at the end of 2018 due to a lack of money. Documents and marketing materials obtained via a recent public records request shed light on why it was unable to attract investors to fund a test track in Colorado to try out its ideas.
In short, it was utterly unclear how their proposal was ever supposed to work, and it remains an example of how messy and unfeasible these transportation startups often are—even as they ask for capital and public money.
The Hyperloop comparison was very rough, since the company’s idea didn’t involve chucking stuff through vacuum tubes. Instead, cars—yes, your own private car, complete with optional soccer-playing child—would be mounted on sleds and propelled through special lanes in highway medians at up to 200 miles per hour using a combination of electric motors and magnetic levitation.
In addition to car-capable sleds, Arrivo also pitched a one-to-eight passenger vehicle and a “Super Metro” people-mover that could transport up to 30 passengers. They claimed the Super Metro could move 72,000 to 108,000 passengers per hour per direction.
For comparison, New York’s L train, which runs from Brooklyn to Manhattan, runs approximately 20 trains per hour during the peak a.m. rush and carries about 24,000 passengers per hour.
Yonah Freemark, a PhD candidate in urban studies at MIT and founder of the influential website The Transport Politic, ran me through the math on this:
In order to transport 108,000 people per hour in 30-passenger vehicles, you would need 3,600 vehicles per hour, or 60 per minute and one every second on average. Given the fact that the average highway lane carries 2,200 vehicles per hour, this would be a 64 percent increase in what is typically understood as feasible on a roadway.
One car a second, that’s all!
Still, it was enough to intrigue the Colorado Department of Transportation, which agreed to a partnership with Arrivo, to build a test track on one of its highways, fund dual feasibility studies, and provide economic incentives through the state’s development arm if they created jobs.
Spoiler: none of this happened because nobody would give Arrivo money to do it and the company shut down at the end of 2018 without ever breaking ground.
Using documents I obtained from a public records request, which I also covered in part for Wired before coming to work at this fine publication, I tried to find out what CDOT could have possibly seen of value in Arrivo’s pitch—other than the obvious allure of a 20-minute-or-less trip to the airport from anywhere in the Denver metro area.
But even after reading through the various marketing materials provided, not to mention talking the government decision-makers, I still couldn’t figure out how exactly Arrivo’s system was supposed to work. And BamBrogan didn’t respond to a request for comment.
The documents I obtained were pitch decks, not engineering studies, but they went into slightly more detail about how the company imagined the system would work.
As with all the documents, Arrivo used a lot of buzzwords without much explanation. But so far as I can tell, the core philosophy behind the Arrivo system is to assume away any points of friction.
Getting onto—and off of—the Arrivo system, which would only run along major highways? “At-speed switching” using what looks like a giant car turntable.
Loading/unloading passengers from the people-movers? An “inventory system” that enables “immediate departures.”
MIT’s Freemark pointed out that, assuming it takes 60 seconds to unload and load each people-mover, the airport would need space, and manage the movement of, 60 vehicles at any given time. Assuming vehicles 40 feet long and five feet wide, the platform would have to be half a mile long, or about 2,700 feet. That’s approximately five times the length of New York City subway platforms.
“This is assuming that it’s possible to clear the platform fully every minute, including acceleration and deceleration, which seems unlikely,” Freemark added. “So you’d probably need another platform. So that’s a total of one mile of platforms at, let’s say, the airport station.”
Drilling down even further, Arrivo provided CDOT a demo video of how the system might work in conjunction with a highway, generally speaking.
A few questions I have after watching this:
- Quickly locking a car, presumably of multiple different makes and models, into the sled for safe, rapid acceleration, not to mention travel up to 200 mph, seems like a huge engineering challenge!
- It’s cool and all how the car-sleds magically re-appear from underground immediately after another car departs, but where are they coming from? Is there a vast underground network of car-sleds? Seems like if you’re going to go through all that trouble to build a vast underground network for moving things, might as well make it for moving people.
- This seems to essentially replicate the on/off ramp system of existing highways, but gaining efficiencies by having the cars on automated sleds. Still, this precisely replicates the problem with merging cars with highway traffic, and it seems like a huge leap of faith to assume computers can efficiently manage a sled speeding up to 200 mph with existing 200 mph vehicles without slowing any of those 200 mph vehicles down. Arrivo assumes away this problem by simply not showing the actual merge in the video. Instead, they simply cut to a section of highway without any on or off ramps.
- I can’t imagine it would be pleasant to sit in any car traveling at 200 mph exposed to the elements other than maybe a top-of-the-line sports car, to say nothing of whether it would withstand such speeds over an extended period of time.
Freemark’s general conclusion was that the entire system doesn’t seem completely impossible, but “it would require very tight spacing and no room for error, I would guess.” And if the Arrivo system can’t accomplish the speeds promised, then it’s difficult to see how it’s any better than a highway system in general.
As far as Colorado taxpayer dollars go, the damage was minimal. Arrivo never built anything, and therefore didn’t qualify for any economic incentive packages. Further, the feasibility study CDOT paid consulting firm AECOM approximately $250,000 to conduct pertains not just to Arrivo but what a regulatory framework for other rapid mobility concepts like Virgin Hyperloop One might look like.
There is a blurry line between innovation and madness. But some ideas are so obviously bad that they can’t survive even the most fervent technocratic fever dreams. Arrivo was one of them, but it surely will not be the last.