I’m in a park. The grass is being mown by a robot.

See the second paragraph. It changes the entire approach to mowing.

(Though I'm headed towards natural lawns that don't need this kind of maintenance at all.)

Consumer electric mowers favor have independent motors for each blade

Big blades are unnecessary with robot mowers. I'm sure there are examples of it, especially in the bigger zero turn mowers that are probably retrofits of existing riding models, but there's actually a simpler approach that works with how automation changes the way it's done.

When you're mowing every day with a robot, you only need to clip a little bit each day. When you do that, you don't need a big heavy blade. Many of those robot mowers have nothing more than a wheel with three or four razor blades screwed on at the edge. They aren't going to hit something heavy, because again, it's doing it every day and keeping everything trimmed all the time. It's safer, too.

When I've talked about my own robot mower at home to friends and neighbors, the thing people have trouble getting over is how you don't just do what you've been doing, only robot. I tell them it trims a small amount every day, and they snicker a little. Then they think about it for a moment and it makes sense.

This cascades down to how the lawn is handled by services. Instead of trying to do the current system, only robot, change the approach. Even if each park isn't going to have its own mower, there can be one truck delivering mowers all over town. Perhaps there are other models that would end up working better, but the point is that swapping out the current equipment for a robot version isn't the way to think about it.

The controller board should scale up without a big jump in cost. A larger zero turn mower doesn't need a significantly different controller to a residential mower. A big cost on those is the GPS sensor. It has to be a relatively high accuracy one for it to work, and that ability has only come down to <$300 in the last few years (it used to be "if you have to ask, you can't afford it" territory). Otherwise, it has to use a boundary wire or something like that, which is the biggest downfall of the one I have. Boundary wires suck.

Once it has enough circuitry to handle the sensors--a larger mower probably needs a few more, but not a huge amount--then it's good enough. Even if it adds $600 to the cost of a residential mower, it may only add $700 to a big zero turn. Cost gets proportionately less as the mower scales up.