Knowledge Brings Fear

Dawn of the robotic age

After lengthy shipping and customs procedures I got finally hold of my first real robotic household item. (In fact not household but company, but that is irrelevant here). It is a iRobot Roombavac automatic floor vacuuming robot. iRobot is a company that sprung from the MIT and is considered one of the leading developers of autonomous systems. They get most of their money from the US Department of Defense for stuff like the military PackBot . The Roombavac is their first consumer product and I must say that it is impressive.

The Roomba is disc-shaped, about 30cm in diameter and 8,5cm high. It runs on a rechargeable battery. The vacuum cleaner is small but efficient and the whole design looks like a typical first generation product where the engineers had still all the pride to withstand the "release now, make it cheaper"-pressure from marketing. The cleaning performance on the wooden floor is ok and the ability to crawl under desks and drawers is really cool. It has some problems with objects that are ramp-shaped, because the bot then drives up the ramp and might get out of floor contact with its driving wheels. The noise is considerable, so letting it run overnight is the best option.

For a first of its kind the Roomba has so surprisingly few drawbacks that it got me thinking about where this trend will go. Currently, house cleaning provides a lot of low-wage jobs that feed a lot of people. In 20 years many of these jobs will be gone. It will start with the lazzy and sociophobic nerds who cannot stand the sight of someone cleaning for them, then move on to the high-security areas where currently only people with the right security clearances can do the cleaning work and will move on to the offices of companies who find it more convenient and cheap to just have one person around to supervise the robots. Households of people with disabilities are also a very likely prime customer group.

The fundamental question is of course how this will fit into the overall picture of society. On one hand, since the younger population in the industrialized countries is declining, there is a clear and obvious demand to find technological ways to enable older people to maintain their household with as little external help as possible. On the other hand, the decline in the number of jobs for people who are currently working on the low end of the wage spectrum is something that can not continue indefinitely without massive social and economic drawbacks.

One thing that might soften the blow a bit is that, from experience with the Roomba, I get the impression that this kind of robots will be in need of a human supervisor for a very long time. Recovering them when they are stuck, emptying the particle bin, repairing the cleaning tools etc. can be done cheaper by a human, at least for some decades. Supervising robots will probably be a very boring job, but might feed some of the "my job went to India and all I got was this fucking T-shirt"-victims.

One very important distinction needs to be made between fully autonomous robots and what Stanislav Lem called "telepresence". Remote controlled vehicles are the majority of things that are called "robots" in the military and search-and-rescue business. Fully autonomous systems and remote controlled vehicles share a lot of development as they both need to solve some fundamental mechanical and energy problems. But not so far down the road the two development path will diverge, as "autonomous" systems require very different sensor setups and will quickly move into physical forms that are difficult to control with the limited sensory and motion set of the human body. As Lem has written a long time ago, the social problems of good working remote controlled robotic bodies ("telepresence") alone can cause some serious headaches.

More detailed information about the current activities in the military field, especially regarding remote controlled and semi-autonomous systems can be found at the Association for Unmanned Vehicle Systems International (AUVSI) .

So in short, robotics is a field that requires a lot of attention, the legal implications alone will make generations of lawyers wealthy beyond my wildest imagination. The economic impact will be, as usual, long term but profound. The first companies will face severe problems, both in terms of acceptance of their product and in the technical field. But it will happen. I have no doubt about that, watching the small disk on the floor, moving purposefully around my shoes, cleaning away the dust.