Tom Ryden on the state of Robotics
This is Thomas Ryden, from Mass Robotics, at an AWS robotics event. I’ll cut to the chase and give you a couple things he says are still open in robotics:
- (Great!) Interoperable, pluggable robot components that work same as PC peripherals (keyboard, mouse, USB). The interoperable robots should be able to co-plan and co-percept.
He says no one is doing this. Question is… how do you find a market for something like this? Does not have to solve the full interoperability problem from day one. But it needs to be a product that is ‘evolvable’ into interoperable robots.
- (Also great!) Provide the cloud infrastructure ‘as a service’ to robots.
Problem: you’d probably compete with AWS eventually.They are building a reference design for this - not an end to end solution, but more a professional services business around this.
What else is in the talk? A great survey of the robotics industry:
- Robot arms (manufacturing).Turns out this is the oldest and most mature segment. Car industry used robot arms for decades. Now, robot arms are becoming intelligent.
- Logistics - warehouse robots. Problem: robots still can’t grip well. Humans are in the loop,picking things and dropping them in a container sitting on top of the warehouse robot, who will move the container to its destination
- Transportation and delivery (ISEE, for example)
- Agriculture
- Construction. This is almost unexplored.
- Inspection, e.g. fly drone to inspect wind turbine
- Hospitality, e.g. robot at hotel front desk
- In the home, helping seniors
Pluggable robots:
- Companies
- Literature
- The Pluggable Distributed Resource Allocator (PDRA): a Middleware for Distributed Computing in Mobile Robotic Networks (2020). “At its core, the proposed approach relies on a task allocation mechanism that maps computing tasks to robots”
- A formal analysis and taxonomy of task allocation in multi-robot systems (2004)