Autonomy vs Intelligence // Studio Day

Today

  • Autonomy vs Intelligence
  • Final Project Studio

For Next Time

Autonomy vs Intelligence

In our discussion about labor and automation, we’ve identified several industries and robotic technologies within them that have morphed the labor landscape. Today we’ll zoom in a bit on the technology itself and examine the computational component to automation.

Recall our definition of a robot as a sensory-motor loop – environmental or contextual data is used to inform actions that the robotic system takes. The mapping between sensor information and actions can be incredibly complicated based on the sensors and actions available, and task specified for a system to complete.

But the complexity of a task does not necessarily mean that the robot is “intelligent” just for completing it. Intelligence refers to the intersection of complexity and ability to self-assign actions or action-sequences. Let’s consider a few different types of systems and map them on complexity and intelligence axes (please feel free to come up with your own additional examples!) –

  • Waste Sorting
  • Manufacturing Arms (for assembly, welding, painting, etc.)
  • Palletizing Robots
  • Warehouse Pick-and-Place Robots
  • Self-Driving Cars
  • Autonomous Trams
  • Combine Harvesting
  • Fruit Harvesting
  • Weeding and Field Maintenance
  • Building/Infrastructure Inspection

After we share out our maps, let’s discuss the impact on markets and labor:

  • What are the trade-offs between complexity and intelligence with respect to:
    • Scalability to a market?
    • Relationships with labor?
    • Resources necessary to create, deploy, and maintain?
  • What market sectors could benefit from intelligent robots? Why?

At the end of this activity, in the place you’ve been keeping your personal notes, reflect on the following question: How does intelligence differ from autonomy, and when is intelligence justified for development?