Use access key #2 to skip to page content.

EScroogeJr (< 20)

A robomower? No way.



January 06, 2007 – Comments (0)

Some people suggested that the next IRBT product could be a robotic lawn mover. There are several reasons that make it highly unlikely.

First, I have already commented on the incredible weakness of IRBT when it comes to mechanical design. Any robotic mower would need better traction than can be provided by plastic wheels a la Roomba. This doesn't seem obvious until you realize the degree to which a conventional mower depends on its human operator to avoid slipping and rolling over. Left to its own devices, a robot must be considerably heavier than an ordinary mover and it must have tracks swimilar to PacBot's (or, at the very least, very robust wheels in the case of very flat and dry terrain). But IRBT doesn't make such wheels or tracks (or blades for that matter). It depends on Deere to manufacture tracks, wheels, and in fact, the entire hardware of PackBot. With this mechanical prowess on IRBT's part, any lawn mover will have to come from a joint project with Deere, which is unlikely to come aboard for fear of cannibalizing its own line of mowers. Or even if it acquiesces, that would come at a price which ordinary consumers, who are less rich than the Department of Defence, will be reluctant to pay.

A partnership with, or acquisition of some small engineering firm might address this issue. But IRBT has made no acquisitions, and has not announced any new partnerships. If IRBT decided to design a mower ih-house, using standard spare parts available in hardware stores, it would need to boost its team of mechanical engineers. I have been reading job advertisements in its site, and found absolutely no sign of expanding mechanical engineering activities. And the current engineering team at IRBT has hardly ever designed anything more sophisticated that plastic cases. The only moving mechanical parts equipped with electrical motor that IRBT seems capable of designing on its own are tiny plastic wheels and maybe brushes. So, the evidence is that IRBT cannot produce a sturdy, robust mechanical device on its own and has not taken any steps to produce it jointly with someone else.

Then, there is the safety part. As a device that has rotating blades, a robotic mower has to include safety features which the Roomba didn't need. As a minimum, IRBT's engineers must implement the following:

a) The robot must stop its blades immediately when it rolls over, or loses traction with the ground.

b) The robot needs an additional sensor that will prevent the blades from hitting a stone. In a dangerous situation, blades must be either stoped instantaneously or pulled inside the device.

c) The robot needs a metal detector so that it can avoid electrical wires.

d) The robot must have a thermal imaging device to avoid any warm-blooded animals such as humans, dogs, and cats (and maybe motion detectords to avoid snakes and lizards), and must stop its blades when in dangerous proximity with an animal.

Now, the request to stop the blades instantaneously when needed means that the whole mechanism must have a very robust design (rotating blades have huge inertia), and is by itself a very serious challenge. Once this difficulty is addressed, part a) should be easy. Now, part b) is already difficult. You need some x-ray system to probe objects underneeth the robot and to distinguish a stone (which should be avoided) from a tuft of grass (which should be cut). Avoiding metal wires is not that hard (look at the price of metal detectors at RadioShack), but I can say with confidence that IRBT does not have any metal detectors in the works. Why? Because Roombas and Scoobas too would sell so much better if they could avoid wires (for instance, computer cables) on your floor, yet no such system was made available even in the latest models. The same goes for thermal imaging systems (which might prevent Roombas from bumping into people and animals). No such feature was ever made available. If IRBT can't even figure out how detect metal, then they cannot figure out how to detect people either. And mind you, part d) is absolutely critical, unless IRBT wants to pay for amputated legs, fingers (what a fine image boost for the company!), dog tails, pet hamsters cut to pieces, etc.

Oh, yes, add to that the ability to recognize water to avoid puddles, and not to attempt "mowing down" wet clay and sand. Sounds easy, right?

Whatever new product IRBT is going to unveil at the CES conference on Mon, I can bet that it won't be a robotic lawn mover.

0 Comments – Post Your Own

Featured Broker Partners