Wednesday, December 26, 2018

Elon Musk's baking soda volcano

As we discussed before in some detail, Elon Musk's proposals can be easily categorized into two groups. The first set are solidly engineered, reasonably viable incremental advances based on standard technological approaches. They are also obviously the work of the many fine scientists and engineers employed by SpaceX and Tesla.

Musk always had a reputation for taking credit for other people's work, and his off script moments over the past few years have made it increasingly clear that the man has a shockingly weak grasp of engineering and cannot possibly contribute to these real projects on anything more than an aesthetics and user experience level. (And yes, that is the same level that Steve Jobs focused on, but Jobs never claimed to be a "real life Tony Stark" despite having a far stronger grasp of the engineering side than what Elon Musk has displayed.)

The second set of ideas apparently do come from Musk himself and they all shared certain common traits. They show a fundamental ignorance of the underlying technology, they are cool and lend themselves to credulous media coverage (it often seems that Musk starts with the CGI animations then throws together a proposal around them), they are orders of magnitude more costly and difficult to implement than Musk suggests and they are invariably derived from familiar postwar science-fiction tropes.

With the debut of the Boring company tunnel, however, we have a new kind of Elon Musk idea, new in a way that raises a number of questions. What reporters saw last week wasn't just a radical departure from the technology that had been promised; it was a completely different kind of Elon Musk proposal. In order to get the full impact of the shift, it's helpful to review what this was supposed to be in the beginning.

The idea was that cars would be carried through a system of underground tunnels using fully automated, high-speed carriers initially termed "sleds." The basic notion isn't all that far from the well-established practice of car-carrying commuter trains, something that might actually be a pretty good idea for certain routes in Los Angeles (it's a very big county and that's not even taking into account the people who drive back and forth from Orange). Between developing the technology, digging the tunnels, and installing the rest of the infrastructure, the cost of this plan would the ludicrously prohibitive, but it would be cool and might even do what it's supposed to in terms of congestion and traffic time.

By going over to the guide-wheels approach that was standard tech in local amusement parks going back to the 50s, the development costs have been largely eliminated, but with them most of the functionality and all of the coolness. The proposed sleds would have been faster and presumably far more maneuverable than individual cars. More importantly, the system would in theory ( and this is all in theory) have been far more reliable. Musk is now proposing a system where any breakdown would freeze up a tunnel for an undetermined period of time. At least based on what we saw last week, removing stalled vehicles from these tunnels is going to be an incredible pain. The result would likely be an even greater tendency toward gridlock than we have now.

Elon Musk has not gotten any better at proposing workable systems, but he still managed to lose his knack for coming up with cool ones. This is a clunky, low-tech "solution" that somehow manages to be both unglamorous and impractical. What happened?

While the following is speculation, it fits the facts and is the pretty much the only explanation I can think of. Musk has had a terrible year and he had locked himself into this demonstration to the extent that backing out would have been a major press fiasco (even worse than what they ended up with). I suspect that at some point fairly recently it became obvious to his team that they would not be able to put together even a crude beta version of the sled, so they opted for the quickest and least sophisticated option, attaching wheels to the sides of a Tesla and simply driving it down the tunnel. The decision appears to have come so late that they didn't even have time to work out a system for retracting the wheels, leaving that instead to the inevitable animation.

Musk is like a school boy who spent the whole semester bragging about the robot he was going to build for the science fair, who then showed up with a project obviously thrown together the night before.

1 comment:

  1. I know it is sexy to point out other countries ability to bore tunnels but this is one in Seattle:

    "The SR 99 Tunnel is a single tube that measures 9,270 feet (2,830 m) long and 52 feet (16 m) wide, carrying a double-decker highway with two lanes in each direction."

    So when a car stalls you can still get access to the tunnel and it runs in both directions. It's also longer.

    I don't love this project, but the idea of a traffic tunnel is old. And these tunnels don;t require guide wheels meaning any car can use them.