The Amazon Robotics Challenge, originally called the Amazon Picking Challenge, was an annual competition run from 2015 to 2019 that tasked university teams with building autonomous robotic systems capable of picking and stowing items from shelves in an Amazon-style warehouse environment. The challenge directly addressed one of the hardest open problems in industrial robotics: unstructured grasping of arbitrary consumer goods with varying weights, shapes, and packaging. Teams used a mix of off-the-shelf robotic arms, custom end-effectors, and novel computer vision systems. The 2016 winner Delft Robotics from the Netherlands demonstrated suction-based grasping with learned object recognition. Cartman from Queensland University of Technology won in 2018 with a novel cable-driven manipulator. The competition accelerated commercial investment in robotic picking and helped establish the field of learned grasp planning.