DoorDash found robotics startups often build tech in a vacuum, then seek a problem. To succeed, DoorDash reversed this: they defined their specific delivery use case (3-5 mile suburban routes) and then designed a custom robot, avoiding the common "technology-first" trap.

For physical AI systems like robots, data quality hinges on diversity, not just quantity. A robot trained to make a bed in one specific lighting condition may fail completely if the lighting changes or the bed is moved. This brittleness highlights a key challenge: training data must capture a wide variety of contexts and edge cases to enable real-world generalization.

As DoorDash's robotics program matures, the primary challenge has shifted. Five years ago, the focus was on making autonomy possible. Today, the harder problems are scaling operations (fleet management, maintenance) and manufacturing reliable hardware at scale.

Dmitri Dolgov explains that while AI advancements create hype, they primarily speed up progress on the initial, easier parts of a problem. They don't change the "long tail" of complex, rare edge cases, which remains the core challenge in achieving full, superhuman autonomy.

Despite testing with countless objects, Ambi Robotics discovered their system struggled with a common item they hadn't prioritized: plastic shipping bags. Bags fold and lose suction unpredictably, highlighting how real-world deployment uncovers critical edge cases that extensive lab testing misses.

Automating science involves solving mundane physical problems. Radical AI had to design custom actuators just to unstick material samples from trays—a task a human does intuitively with a chisel, highlighting the often-overlooked 'last-mile' challenges in robotics.

Self-driving cars, a 20-year journey so far, are relatively simple robots: metal boxes on 2D surfaces designed *not* to touch things. General-purpose robots operate in complex 3D environments with the primary goal of *touching* and manipulating objects. This highlights the immense, often underestimated, physical and algorithmic challenges facing robotics.

Waymo robo-taxis are calling 911 because passengers are falling asleep. This isn't a critical system failure, but a human-robot interaction problem. It reveals that successful automation requires solving not just complex technical challenges but also simple, unpredictable human edge cases that arise in real-world deployment.

Moving a robot from a lab demo to a commercial system reveals that AI is just one component. Success depends heavily on traditional engineering for sensor calibration, arm accuracy, system speed, and reliability. These unglamorous details are critical for performance in the real world.