For a long time now, there has been talks of a robotic revolution which seems to be continually pushed back to the next day. Robot vacuum cleaners, IOT, the autonomous cars are the beginnings of this change. But why is it so slow?
On one hand there are living beings, shaped by reality, with efficient mechanisms, which are capable of managing and exploiting their environment. On the other hand there are the machines, shaped on the virtual, for which reality is extremely hard to master. This duality is the main issue.
The automation of the virtual world that the computer allows us to do is no longer enough for Humanity; we must automate reality and robotics is essential to our evolution.
Now, the technologies are ready and this revolution is imminent. In the coming decade, the costs of these technologies will drop to the point where they become accessible to the general public, and we will witness a frantic race of robotics companies with huge growth, as we have experienced with computers in the 80s.
The companies that will succeed in developing always better, always faster, and always cheaper will be the ones that will win the race.
Robots are very complex systems, made up of a large variety of autonomous sub-assemblies, the intelligence of which must be coordinated and pooled into coherent behaviors.
To improve their performance, companies will have to abandon the current vertical development model, and encourage a new horizontal and collaborative development. In order to do that, it will be necessary to reuse and outsource some expertises to companies that will specialize on some sub-assemblies. Modularity is essential. Indeed, these companies need to use common technological tools making it possible to easily share and perpetuate their developments.
The world needs a technology inspired by the nervous system of living beings to coordinate all the elements of the machine efficiently, safely and economically into a single system image.
We need a technology that can group the robots' functionalities into standard and easy-to-use coding objects. It would provide isolation between software and hardware and it would smooth the transition from a prototype to an industrial version. It would also make it possible to optimize the apportionment of processes in the various sub-assemblies, and to relocate some costing algorithms, for example in the cloud.
Robotics also have to be able to adapt dynamically to the addition or removal of sub-assemblies in a robot, allowing to actively change the position of an element (like tools for example), because this is essential for the evolution and maintenance of the machine.
Manufacturers can also offer basic, simple, low-cost versions, with the possibility of adding options like what we see in cars today.
Such a technology needs to be created for machines that will be produced in large volumes and which have quite large dimensions and cost constraints.
And this is exactly what we're working on at Luos!
Thanks for reading! The Luos' Team