GO, Waymo and Nihon Kotsu: What a Three-Way Robotaxi Partnership Tells You About Who Owns the Rider

Waymo supplies the autonomous driver — the hardest, most capital-intensive layer, and the one with the least local advantage. Nihon Kotsu supplies vehicles, depots, maintenance and decades of Tokyo operating licence history. GO supplies coordination and, critically, an app with 35 million downloads and roughly 70% of monthly active taxi app users.

Each party protects its own layer. None of them has an incentive to build a surface that also sells a competitor's rides.

A visitor landing at Haneda does not think in partnership structures. They think: how do I get a robotaxi in Tokyo. Today that question routes them to a coin-flip between GO, Uber, S.Ride, Didi Mobility Japan, or a hotel concierge.

The fourth role is the neutral aggregator — the layer that takes intent, matches it to whichever fleet can actually serve it, handles identity, language and payment, and clips a small transaction fee. It owns no vehicles and licenses no autonomy stack.

GO's value to Nihon Kotsu is that it drives demand to Nihon Kotsu vehicles. The moment GO becomes genuinely fleet-agnostic it competes with its own partners. Uber has the same constraint with Wayve and Nissan. Neutrality is not a feature you bolt on; it is a founding position or it is nothing.

That is why aggregation layers in every other transport category — flights, rail, hotels — were built by outsiders, not by the operators.

A neutral layer only works if riders can find it without being told. In practice that means owning the words people already use. In Tokyo those words are robotaxi and Tokyo.

robotaxi.tokyo is the exact-match version of that intent. We are an early-stage group building the thesis, not a live service — and the asset is open to acquisition or a seed partnership with someone who intends to build the fourth role properly.