• Start Time:
    12:15PM
  • End Time:
    12:45PM
  • Day:
    Day 2

Talk:

The US has labor shortages in warehouses and manufacturing.  For such indoor environments, even with seasonal human workers present, it is now possible to deploy AI-based robots for whom there is no long tail of unexpected situations. Not only are the fruits of ML directly useful, but the impact ML has had on silicon architectures gives us a double bounty for more traditional spatial perception, awareness, and reasoning.  All robots in a facility contribute to enabling robots to fully understand what is happening everywhere in a facility.

By making robots human-centric first, the robots can increase both productivity and job satisfaction among hard-to-retain workers. Such robots not only offload the grunt work of moving physical objects within the facility but through human-centred interaction design the robots can offload much of the cognitive load from people by giving them cues on both where to direct their attention and monitoring task completion, enabling them to be much more productive in their work.

  • Vision-based SLAM outperforms LIDAR-based positioning and lets the robots know at all times exactly where they are in a shared map.
  • Through mixed-initiative and cognitive support for workers our trials have shown productivity improvements of 30% to 40%.
  • By being just the right amount of deferential to human workers the robots are not seen as either a safety or employment threat, and instead become valuable tools.

Associated Speakers:

Rodney Brooks

Professor Emeritus

MIT

Associated Talks:

12:15PM - Day 2

View Presentation: AI-based robots that complement human workers in warehouse and factory floors, radically increasing efficiency

View Full Info