radical Insights.

Weekly Research and Commentary on the Future of Business and Technology.

Don’t Buy Into the Internet’s Exponential BS Machine.

Jan 15, 2019

The Internet buzz machine is going bonkers over the latest “breakthrough” in robotics: Autonomous Delivery Robots. From Postmates’ bug-eyed Serve robot, and Starship’s little six-wheeled cooler-on-wheels look-a-like, PepsiCo’s fleet of snack-delivering bots, to stair-climbing delivery machines making waves at CES. Around the corner from radical HQ we even had our first spontaneous combustion of a delivery robot on the UC Berkeley campus.

The future of last-mile delivery in cities is clearly in the cute, multi-wheeled mighty hands of little autonomous delivery robots, right?

No.

As so often with new tech – I believe the current approach is destined to fail.

The first time I saw a Starship bot at Singularity University a couple of years ago my initial thought was: I wonder how many of these things will be vandalized and ditched into the next river (joining the hordes of bike-sharing bicycles and electric scooters)? And how much fun must it be for a teenager to block the way of the next pizza-delivering robot? Just stand in its way and wait for the pizza to get cold and the delivery is late. Moreover, have you been to a city center lately? The whole point about these delivery robots is that they share the pavement with pedestrians (so that they don’t need to maneuver alongside cars and other vehicles). The last time I was in London, Berlin, New York or San Francisco the pavement is busy – as in really busy. Hard enough for a father or mother to maneuver with a stroller – quite the feat for an autonomous robot to move through (and make some headway instead of simply being stuck for hours on end).

That all being said, there surely are use cases for these delivery robots, where they make perfect sense. Closed environments, factory floors, etc. – but delivering pizzas in New York? Nope.

Which highlights the classic tech fallacy: Just because you can, doesn’t mean you should. Focus on the customer benefit, take real-world constraints into account and build what matters from there.