Coco Robotics is moving its delivery robots beyond the sidewalk. The company says its new Coco 2 bot can operate without remote human drivers and can use streets and bike lanes at speeds of up to 13 mph.
Niantic Spatial, spun out from the company behind Pokémon GO, is part of the navigation system behind that push. Its mapping and visual-positioning tools are now being used to help delivery robots handle city routes where GPS can drift, and drop-off points are harder to pin down.
Coco is pushing beyond the slower sidewalk delivery bots that have shaped the category so far. The new robot can move onto streets and bike lanes where legal and appropriate, and it can carry multiple orders at once, according to Semafor’s report on Coco 2.
Niantic’s mapping
In dense city delivery, knowing the block is not enough. A robot has to stop on the correct side of the street, approach the right entrance, and handle curbs, bike lanes, loading zones, and tall buildings that can throw off satellite positioning. That is the kind of navigation problem Niantic Spatial is trying to solve.
In Niantic Spatial’s partnership announcement with Coco Robotics, the company said Coco will use its Visual Positioning System to improve location accuracy in dense urban areas where GPS can be unreliable. Niantic Spatial also says its geospatial AI platform is built on more than 30 billion posed images from millions of locations, creating a large visual base for machines moving through the physical world.
Niantic built its name on Pokémon GO, the game that sent millions of people out into streets, parks, storefronts, and landmarks with their phones in hand. Now that same spatial foundation is being used in robotics to help machines stop at the right curb and navigate city streets with more precision.
What Coco is trying to prove
Coco says it has completed more than 500,000 deliveries and operates 1,000 robots across the US and Europe. In Coco’s launch announcement for Coco 2, the company said the new model is designed to adapt more quickly to new cities and improve through real-world operating data.
Street and bike-lane operation adds a harder set of conditions than sidewalk delivery. Robots have to share space with cyclists, deal with tighter safety expectations, and navigate local regulations with less room for error.
Coco says the robots were trained on conditions including flooding in Miami, snow and freezing temperatures in Chicago, and dense traffic in Los Angeles.
Whether cities embrace that more expansive model is still an open question, but Coco is clearly trying to move delivery robots out of the novelty lane and into everyday urban logistics.
Also read: Amazon’s Blue Jay robot stalled within months of launch.
