Every industry will experience increasing pressure to collect data from, and maniuplate, the physical world: e-commerce, food, advertising, sales, manufacturing, agriculture, healthcare, labor management. This will require us to understand sensors and autonomy. We hear that making hardware is hard, but what if hardware isn’t hard? What if lean enterprise ideas worked just as well for hardware as they do for software, maybe even better?
When we accept this idea, we have the freedom to embrace failure and to release imperfect versions. This gives us the freedom to iterate like crazy - not only do we learn a ton about the problem space, we also grow our skills along the way.
Any device that will survive first contact with the market needs about 15 testable iterations before we really discover pmf, manufacturability and durability. So multiply the cycle time by 15 to get a good estimate of when we can be in the market. We really want to get this down to a week per releasable iteration, and there are ways to do this.