Minecraft pic.twitter.com/lU1YzJjLOZ

— Elon Musk (@elonmusk) February 4, 2017

The tech leader first started publicly musing about the project in December as a way to address urban traffic congestion. In typical Musk drollery, he said he would drill tunnels under the auspices of “The Boring Company.”

Get Data Sheet, Fortune’s technology newsletter.

At the time, he had to reassure observers that he was entirely serious. But he officially started work last weekend, despite saying in late January that it would take “a month or so.”

Being ahead of schedule seems to be part of the project’s broader agenda. Musk told Wired that the goal is to improve tunneling speed “somewhere between 500 and 1,000%,” and that the starting point was a “test trench” on the grounds of SpaceX’s Hawthorne, California headquarters.

It’s interesting that Musk has taken a more personal stake in this tunnel project than he ever did in his other big transit idea, the Hyperloop. Tunneling may certainly seem like a more low-key project for a man who already heads two huge, high-risk companies—but the reality is that transit-oriented tunnel projects are often much more challenging than just drilling a big hole.

Just one recent example was the plight of Seattle’s Alaskan Way Viaduct replacement project. The massive drill used in that project, dubbed Big Bertha, hit an unmarked well casing only 1,000 feet into its route, then sat immobile beneath the city for the better part of two years. That project also threatened historic buildings over the planned tunnel route.

There are similar hurdles to tunneling, above and below ground, in all the urban areas that would most benefit from it. So while faster (and thus cheaper) tunneling could definitely make going underground a more appealing option for future city planners, cost isn’t the only obstacle.