Elon Musk announced early in development of the Model 3 that Tesla would be able to produce 5,000 cars per week due to their use of “hyperautomation.” Although robotics had already revolutionized the assembly line for other car manufacturers, Tesla was adding computer vision to their assembly line that allowed robots to “see” and react to unexpected car variances in real time using artificial intelligence without human intervention. Musk purchased hundreds of robots more than any other car manufacturer, and when he saw a bottleneck, he installed more robots to smooth it out. He bragged that he was building the factory of the future, which would be a key competitive advantage in the future.

We all know that this was a huge source of problems for Tesla during the early days of the Model 3 (“production hell”), requiring the hiring of thousands of humans to help with assembly. A disappointing 2,000 cars were produced a week, far below his original estimates, and far below the 6,000 cars a day that the 5-million-square-foot plant produced for its previous owner, Toyota. The delays were so great that it almost bankrupted the company. For the most part, the kinks in the robotics software algorithms have been worked out, and the balance between humans and the 160 robots in the assembly line tweaked, showing that Musk had the right idea, it just took the world of robotics a few years to catch up to his vision.

Amazon also went full tilt into robotics, believing that they would be much faster than humans at boxing and shipping products in their massive distribution warehouses. Jeff Bezos was wrong. Humans are still much faster at finding and boxing products. Unlike humans, robots have one speed. Humans are easier to flex in times of increased need around holidays. They pack more efficiently and faster than any robot. Both companies also failed to ensure that appropriate preventive maintenance was performed on their robots, pushing them to their limits and then, when issues arose, using temporary fixes instead of the proper repairs.

In this issue, we discuss the use of robotics in retinal surgery. The idea employed here is what Musk and Bezos quickly learned, which is to use robotics to improve the efficiency of humans, not to replace them. Humans can adapt to issues immediately; in retinal surgery, this can mean the difference between vision or permanent blindness. The actions and algorithms in these devices still need a lot of work. Certainly, the use of robotics is in its infancy, but as with many things that we now see as routine, we must start somewhere. We also look at all the areas where technology is improving our efficiency as retina specialists. But like the giants of industry at Tesla and Amazon, we need to be careful how we use technology and not make the same mistakes they made. RP