Skip to main content

A real hyperloop is almost here — and it’s not what Elon Musk envisioned

A real hyperloop is almost here — and it’s not what Elon Musk envisioned


It’s slower, bigger, and less loopy than Musk’s initial spec

Share this story


This morning, a company called Hyperloop One made history. Nearly four years to the day after Elon Musk first published his plans for the system, the company set a new hyperloop speed record, logging nearly 200 mph on their test track outside Las Vegas. It’s an impressive achievement by any measure, but it’s particularly impressive given how implausible the project originally seemed. When Elon Musk first announced his plans, few observers would have predicted that a workable prototype would emerge within four years — particularly one developed without Musk’s direct involvement.

Hyperloop One still faces significant challenges as an actual business venture, but it’s worth appreciating how much it’s achieved so far. The company has successfully brought hyperloop technology from a woolly thought experiment into a viable prototype that can do almost everything Musk described four years ago.


If you compare the existing systems against Musk’s original specifications, you can see a number of compromises. Each one is understandable, maybe even inevitable, but they add up to a significant change from what Musk originally proposed. As the hyperloop has come down to earth, it’s lost some of its luster, becoming more like a faster, lighter train. And in the process, it’s raised serious questions about how much of an advantage it will have over existing transit methods.

It’s the size of a train

The initial craft envisioned by Musk was the size of a bobsled, carrying three passengers in single file with almost no room to spare. This was a crucial element of the spec, allowing the tube to lay over a series of concrete pylons with minimal footprint, hopping over mountains, and even running up the median of I-10. This was always a long shot, but it let Musk plot a straighter course, imagining hyperloops where a highway or railway track would be impossible.

Hyperloop One’s design has had to scale back that idea. With the tubes at 11 feet in diameter, it’s closer in size to the cargo version laid out in the initial Musk spec, along the lines of a small train car or bus. The track used in today’s test is on simple embankments, and while Hyperloop One imagines it eventually being elevated on top of pylons, they’ll be far larger than the narrow columns initially proposed by Musk.

It’s easy to understand why Hyperloop One made the shift. It would be difficult to load a small three-person craft efficiently, and with all the passengers headed to the same destination, there’s little to be gained by transporting separate crafts. The simple claustrophobia of a coffin-sized shuttle might have been too much for some passengers. Still, the result is a transportation system roughly on the scale of a train or highway, without the disruptive potential of a smaller and lighter tube.

It’s not a loop

It’s still hard to say exactly what Hyperloop One’s final product will look like, but it seems more and more linear. Early layouts indicate multiple branching routes, which would be incompatible with the unidirectional loop initially laid out by Musk unless separate tubes are built for each destination. Today’s test focused on the air-rail and propulsion systems, forgoing the more complex ventilation systems that make the loop system worthwhile. If that pattern holds, the result will be that the hyperloop simply isn’t a loop anymore.

This seems like a pedantic point, but it has real implications for the overall efficiency of the system. Musk’s initial spec was self-powering, counting on constant high-speed winds in a single direction through the loop, speeding the craft in the early stages and being driven by the craft later on. That system requires a complex on-boarding / off-boarding system that’s largely missing from Hyperloop One’s plans, which more closely resemble a traditional train.

The claim that the hyperloop would be self-powering was always something of a fudge, counting on miles of track-mounted solar panels to provide a steady power source. But the broader idea was new and compelling: creating a closed system with as little friction as possible, and a strong, self-sustaining current in a single direction.

That idea isn’t present in the current generation of hyperloops, and it’s genuinely unclear whether it will ever reappear. Without it, the hyperloop takes a small but significant step toward just being a lighter maglev. It also makes it more difficult to get up to speed, which brings us to an even bigger problem...

We still don’t have a good sense of top speeds

How fast can this thing really go? This morning’s test topped out at 192 mph, well short of the promised 250 mph, even though there’s little doubt that Hyperloop One will eventually hit that mark. (For comparison’s sake: Japan’s bullet train routinely reaches 200 mph, and maglev tests have hit speeds as high as 350 mph.)

The initial hyperloop plans were a lot faster. In Musk’s initial announcement, he presented the hyperloop as a system that could get you from Los Angeles to San Francisco in less than 30 minutes, which would require an average speed around 700 mph. Throughout the spec, Musk assumes the craft will travel between 700 and 760 mph, which is more than twice the speed of existing maglevs or the Hyperloop One tests. The ability to reach those speeds was the biggest single factor setting the hyperloop apart from conventional rail travel — and it’s still unclear whether the hyperloops we have will be able to close the gap.

From an engineering perspective, there is a huge difference between 300 mph and 700 mph, so much so that Musk classified acceleration above 300 mph as an entirely separate propulsion issue in the initial spec. The higher speeds come with difficult G-force issues, requiring straighter paths that make right-of-way problems even harder to solve. Safety problems become more serious, too, raising the stakes for even the smallest failure in the track or craft. There may be ways to solve these problems, but so far, none of the existing hyperloop groups have even addressed them.

For now, Hyperloop One is punting on the speed question, for understandable reasons. The company has said it wants the crafts to go as fast as possible, provided it’s safe, but press announcements have stayed clear of any Musk-esque pronouncements on travel time. With only 500 meters of test track, pushing the 700 mph limit won’t be practical for a while, and there’s nothing in the company’s near future that requires a 700 mph Hyperloop rather than a 300 mph one. The slower version is simpler, easier, and it already works. Why make things difficult?

In some ways, this was unavoidable. Any engineering project requires compromises, and taking on something as ambitious as the hyperloop requires more than most. But it’s also a reminder of how much earth-shaking ideas can lose in translation. Like Tesla and SpaceX before it, the existing hyperloop is an incredible achievement — but it’s not quite the world-changing idea it seemed to be in 2013. The initial designs had obvious benefits over existing tech: they were faster, cheaper, and more efficient than alternatives like high-speed rails. Now, those benefits are less clear. It raises a tricky question for Hyperloop One and its competitors: when the finished version finally does appear, will it be worth the effort?