iRobot Roomba Review: Part II

Day Two: Part II continuing from earlier post.

Did it clean: I woke up to a freshly vacuumed carpet. Of course the carpet was already clean, it cleaned it the day before, but our Roomba had come and vacuumed in the middle of the night leaving only two signs that it had been there—the vacuum trails left behind and the fact that it was sitting there in front of the door. When I went to replace him on the dock I noticed that the dock was twisted to the side so that the robot couldn’t find it’s base. I figured that the Roomba must have bumped the dock and twisted it, but it seems to avoid the dock (like the virtual walls) while cleaning except for one time… when it’s docking. I tested it a couple times and finally got Guster to duplicate what it did to twist the dock. Incidentally, it twisted the dock again on the morning of day three as well, but it was still able to dock. I think that I’ll have to find a way to  ground the dock so that it won’t get bumped in the middle of the night. See it in the video below. As it is, it seems to correct itself most of the time (as it did in the video) but it won’t always.

Cleaning the Roomba: I then open Guster’s  guts to reveil what he’d had for breakfast (yes, we call him Guster, after Gus on Psych). It seems he likes a mixture of dirt, crumbs, and small bits of paper (in short all the things I’d rather not eat minus the olivs). Emptying out the bin and cleaning the filter were again revealing. Even though he had only vacuumed a scant fifteen hours before there was a good bit of debris (this sheds light to the vacuum sales fallacy: that you vacuum with your old vacuum and then they vacuum with theirs and you see how much yours miss, well next time vacuum with yours again and see how much theirs miss) of course I expected this to some degree. It picked up about a third as much as it had the day before.

The brushes were much easier to clean the second time around. On day one, it took me about 3 minutes to clean (once I got the hang of it) using the handy circular comb that came with our Roomba and a steak knife (to clean around the bearings, couldn’t do it without a knife). Today it was far easier. There was hardly any need to clean the brush. The paddle on the other hand (the brushes counterpart) had some yarn wrapped around one of the bearings which had collected dust and hair. There was some dirt and hair around the other side too, but not nearly as much. Today it took me about two minuets to get everything clean and put back together. Much faster than vacuuming.

I actually wouldn’t say that the Roomba is too loud. It vacuums away while we slumber peacefully in our bed. It seems quieter than a normal vacuum and it still seem quite effective—though it still sounds like a vacuum (shame on the Roomba people).

[/caption]

Day Three Through Seven: We have woken to clean floors every morning this week. It has found its dock every time except one giving it a 85% success ratio. We’ve had it clean our bedroom and other rooms in our house while we did grocery shopping one day—it even cleaned under the bed for us (which we never do). In fact, I would say that Guster cleans better than we do.

What would I change in the Roomba: I love that my floors clean themselves, nevertheless there are a few things I would change with my Roomba.

First, iRobot does not produce a HEPA filter. Personally, I think this should be standard. If you check their FAQ, it will say, “You can schedule Roomba to run when you’re out or in another part of your home, thereby eliminating allergen inhalation during operation.” Yes, that’s true, but that’s besides the point. Not only do allergens linger for a while in the air after the vacuuming is done, those same particles will settle on carpet and furniture and will be stirred up by walking over the carpet, picking up objects in the room, and the next vacuuming. There is another option, RobotAddOns makes a HEPA filter for Roomba. While it might be considered a breach of your warranty, it is still an option.

Second, I’m not sure that the random algorithm used to send our Roomba off on various tangents is the most effective way to clean. It seems that the vacuum estimates the size of the room, sets a timer, and returns to its dock when the timer is through. there are a few other companies (none so popular or as convenient as the Roomba) that are producing cleaning robots. While I haven’t worked on the problem myself, the solution seems to be simpler than all of these companies make it. One company creates a “North Star” infrared pattern on the ceiling where the robot is cleaning for a reference point. Another company uses markers (which looks like a kind of tape) to mark off the areas where the robot may clean. While both of these other methods enable the robot to vacuum in a line (and thus shortening the vacuuming time by limiting how many times it vacuum the same places) I think that both methods are unnecessary. Can a vacuum not be programmed to measure out distances? If so then a robot can determine when it has found its original starting point after crawling along a wall. And if it can do that, then it can move inward around the room until it is finished, still avoiding and moving around obstacles, etc. Maybe I’m wrong. Let me know if you see any problems in this.