Autonomous robots in a 2016 world

This BloombergBusiness article on Google’s plans for robotics is an example of the pessimism with which industry and companies see the future of commercialized robotics and the financial rationale behind it. After Google acquired Boston Dynamics in 2013 as part of a spree of acquisitions in the field of robotics, now puts Boston Dynamics for sale.

Google wants practical solutions that produce revenue for the company. As a software company understands that software is far more easier to develop than hardware and cannot see how the impressive robots of Boston Dynamics can produce profit in the next few years. The huge scientific problems in both robotics and AI can be tackled with high-risk fundamental research, and Google seems to prefer to invest mainly on AI. Unfortunately, this is indeed the reality: software is far more predictable to develop and the theoritical research behind it far more easier to be implemented. With hardware and real physical systems the problems are bigger, sometimes even the theory fails to fully predict every aspect of a physical environment.

Having a humanoid robot interacting with a dynamical and chaotic environment robustly and safely is beyond the current scientific state of the art. In fact, we are decades behind this reality and this is something that the head of Boston Dynamics profoundly understands. That is why Boston Dynamics division refuses to work with the rest of the Google’s roboticists. Google wants to push the division on the right of the Research and Devepment axis*, i.e. towards development.

In conclusion, the sale sign verifies the fact that robotics needs decades in order to solve real-world problems, as even Google (which has spend a lot of money on high-risk projects) gives up on the highly advanced robots of Boston Dynamics.

*In the far left lives the scientific fundamental research, with high-risk for investment and long-term results, and in the far right the technological development of ready-to-use products.