Nav

Rod Brooks interview: TLH notes from the Nantucket conference

MIT professor, star of movie Fast, Cheap, and Out of Control

iRobot started without a plan, just some JPL contracts, ran that way for 8 years

Did not raise money until we saw external market opportunities

• At one point had six product divisions
• Now two divisions:  gov and consumer
• If we had it to do again, we would work harder to figure out the market in advance, and do less trial and error
• Biggest gov product is the “pack bot” — 50 lb unit
  
  • 2,000 shippped
  • Used by troops to handle IEDs
  • 19 year old kids with little training
  • very rugged
    
• Building a 30 lb pack bot to be issued to each squad in 15-50 brigades
• Will make gov the dominant part of the co
  

Sold Roomba using late night TV, looked for fast results (ads pay for themselves)

• 4m sold
  

Got the key technological insight from insects:  have 100k neurons, navigate reactively, cover ground fast, don’t try to build a detailed model of the world [advice we should all take to heart]

HeartLand robotics

• “low cost robots to empower American workers”
• few industrial robots today, most of them based on Iron Hand design, good at following a complex trajectory repeatedly and precisely, use-cases designed around this ability, integration cost = 10x capital cost => mostly used in cars
• bringing the iRobot design philosophy to industrial robots
  

See Robotics as a way to both keep manufacturing in US and address the labor shortage that will occur as the boomers retire

• Unions realize that American workers need to keep getting more productive to keep jobs — to keep entire factories from migrating overseas
  

Korea putting a lot of investment into robotics:  focusing on functional robots like Roomba

• Japanese are “lost” with their humanoid robots, which are frequently tele-operated
• US has leadership in robots for unstructured environments, driven by the military, aided by consumer adoption