Here’s a little of what I’ve been working on at Google X for the last few years: generalized robotics. To my understanding, we have built the largest fleet of mobile robots capable of interacting on the human scale (ie. Picking up things, open drawers and doors, wiping tables). There’s still a ridiculous way to go in this field! I’m finding the most important thing in navigating all the possible solutions and technologies is to understand the problem
Generalized Robots: as a ‘helper’ robot it has to navigate unstructured and cluttered environments. As opposed to most robots currently out in the world being choreographed, remote controlled, or with explicitly coded reactions. For example in Mechanical Design, it’s “how big or strong or small or fast does the gripper need to be?”, “Should the cameras be high up to see all around or lower to see under tables, or on the arm for different vantage points?”. Ultimately these all hit multidimensional tradeoffs, so only when understanding the value and the problem can we dive into all the crazy solutions and technologies. There’s still a lot to reflect on, happy to chat and hear other thoughts.