I used to work at iRobot. Neato doesn't understand the problem.
Consumers don't care how the robot works. They just care that it does. The whole point of robotics is to not care about the problem. Who cares if there is a map?
I also don't think people can say a roomba is dumb. It is designed with few sensors, but the software under the hood is actually why it is good. It uses lisp in a subsumption architecture, compiled down to run on a dirt cheap processor. Most companies couldn't compete with this on a cost basis if they tried. Here we see more CPU and sensors thrown at the problem, which is much more expensive.
I'm a consumer of cleaning robots, and I certainly care.
I'm on my second Roomba now, and I love the thing to death.. However.. The fact that it bumps around randomly for a very long time while missing obvious spots can truly suck. Your Roomba can go several whole cleaning missions without ever touching a given spot, or mostly ignore half your floorplan for whatever reason (i.e. the shape of the room). Neato appears to be planning to sell this thing for $399, which is around the same price range as a Roomba. If there turns out to be a price premium when this thing hits the market, it will be well worth it if it turns out this thing works better due to the allegedly-expensive software and hardware.
And that's the whole point -- what it comes down to is not nerd-love or whatever, but that it really works. Every single person that I've ever met and discussed my Roomba with asked the same first question: "does it actually work? does it map out your room? no? so how does it make sure that it hits every spot?".
I can reply with iRobot's hopeful marketing or I can tell the truth: it works pretty well, but not as well as you'd expect. A device that would actually ensure every spot is visited once, whatever the method, is going to be a winner, all other things equal.
I think consumers, geeky or not, have the capacity to both care that a device works and spot obvious methodological drawbacks.
I have a Roomba. It simply doesn't work well. There are always spots that are missed in a medium sized room even when I leave it running for half an hour. I just end up vacuuming the place myself since it takes about a minute to get 95% coverage instead of 75% coverage in 30 minutes.
I owned a Roomba and set it to run while I was at work. I would never see it operate. I just emptied it every day.
However the chaotic way in which it patrols does mean it cannot replace a conventional vacuum for the roles of spot cleaning or for a touch-up just before guests arrive. It completely falls down in that role. Even using its spot cleaning mode it would miss a fair amount of obvious debris even if it was placed right on the most concentrated areas. And for touch-up before a guest arrives? You better have a conventional vacuum. Sometimes it would be unable to exit rooms even with the guidance beacons. I would circle around and never be able to get out of a conventional doorway. I'd find it dead in one of the two rooms I wanted it to do for the day, and found evidence that it didn't tackle a room at all. About once a week. And this was a 5th generation model.
To me, the Roomba was to complement your existing upright or canister vacuum and was incapable of replacing it entirely. There would be greater success for a product that could replace a conventional vacuum entirely.
I have to get one for my grandparents. They are the type that will vacuum a room and walk out backwards while vacuuming just to get the straight lines in the carpet. The Roomba doesn't leave the straight lines, and it bothers them, so they don't use it.
So yes, while they won't care how it works, they will like that it vacuums like they do. And that means something.
There are certainly a number of people who would buy this for that reason. But I think of the people who are particular about vacuuming, many more of them won't buy a robot at all. I can't imagine selling millions of this because of people who care enough to have straight lines but not enough to do it themselves.
The scale really does matter too. Software and hardware are very different businesses. If they don't get to a sufficient scale, their cost per unit will be higher, and their margins much smaller.
While I agree with others responding to this comment, my major complaint with my Roomba and Scooba is the ongoing maintenance. If I don't thoroughly disassemble and clean the robots after every couple uses, they will stop working. Hair and fibers end up wound around the brushes and wheels and quickly stop them from working entirely. My vacuum does not have this issue. My broom does not have this issue. My mop does not have this issue. Honestly, it feels like I spend as much time cleaning my robots as it would take to just vacuum and mop by hand.
the dirt cheap processor is not reflected in the price of a roomba. i enjoy mine, but it has issues still (and it is the latest rev) that smell of vacuum 1.0. I hope the neato is a step towards 2.0. As a side, I think this is great. Competition keeps innovation rolling.
It says the Neato avoids humans, so I'm assuming it also avoids pets. So now I wonder how much work the Neato will get done around a territorial house rabbit. I think I'd feel sorry for the robot, it's either not going to be able to clean one of the dirtiest sections of floor (sawdust and hay) or it's going to have to keep deploying to its own personal Iraq. I'm glad to see there's no exposed rubber that can be easily stripped off.
I know the Roomba wouldn't have survived my rabbit, too many raised plastic edges and no ability to avoid him. I would be interested in seeing if the Neato would be capable of entering my rabbits territory over concurrent visits. Although I'd definitely have nightmares of the rabbit hijacking the Neato.
On the upside, it gave me a chance to take the thing apart completely. Roombas are very meticulously designed and are built for being taken apart. It matches up completely with their policy for allowing and encouraging modification of the firmware.
I just wish I didn't have to deal with feces to learn that :S
If they use a litter box, add some sand to it. For some strange reason, if they shit in sand they'll never shit anywhere else. Maybe it's some remote gene from back when they existed solely in deserts, but I know my two cats repeatedly ruined my sandpit as a child. You've literally got to replace all the sand and wash the pit out if you want them to stop.
I don't think the list of 'features' on the neato are the type that the average person would understand/care about. They both clean rooms, they look fairly similar, they both know how to charge themselves, etc. etc.
I think neato needs to differentiate themselves through some other means. Roomba has the mind-share and brand name (though neato is a good name).
The cats will certainly run for their lives, just like when I pull out the vacuum cleaner. I'm sure they will adapt and it's better not to be associated with the mean thing. However, does it do fringes? I heard that the Roomba does not and the oriental carpets have fringes.
I just feel like these products just aren't there yet. If I showed this video to my wife the first thing she would say is that it missed 1" around the wall and it didn't do under the couch.
For someone living in the city in a small apartment this really isn't worth it yet.
Consumers don't care how the robot works. They just care that it does. The whole point of robotics is to not care about the problem. Who cares if there is a map?
I also don't think people can say a roomba is dumb. It is designed with few sensors, but the software under the hood is actually why it is good. It uses lisp in a subsumption architecture, compiled down to run on a dirt cheap processor. Most companies couldn't compete with this on a cost basis if they tried. Here we see more CPU and sensors thrown at the problem, which is much more expensive.