IClebo Arte and Neato XV-21 vacuum cleaning robots - comparison based on experience
As a lover of any kind of automation and automation, I have long been thinking about acquiring such a miracle device, like a robot vacuum cleaner. On the market, these devices have been around for a long time, but until recently their level of “intelligence” remained too insignificant for me to be recognized as robots. And in the budget segment, really "smart" models finally began to appear. The experience of practical use and the final choice of one of two such models will be discussed in this article.
Here I will try not to focus on the characteristics, the review of structures and other facts that can be gleaned from numerous descriptions, official and not so much. I mainly want to share my observations on the behavior of two completely different robots, on the effectiveness of their work and ease of operation. That is, the information that I could not get in any other way than a personal test drive, and which I did not have enough to make a final decision. Perhaps someone this article will help determine the choice.
So, after studying the specifications, prices, the possibility of buying, reading reviews and watching the video, only two vacuum cleaners remained in my attention - the iClebo Arte produced by the South Korean company Yujin Robot and the Neato XV-21 from the North American company Neato Robotics ". The intelligence of both of these models is implemented on the basis of an embedded Linux computer. The first one in my house was iClebo Arte, I used it for about a month. Then I watched a rental Neato XV-11 for two weeks, after which I purchased the Neato XV-21, which I have been using for a week at the time of this writing.
')
The main characteristic that allows, in my opinion, to recognize robots vacuum cleaners as just robots, is fully automatic intelligent navigation. And it was the logic of constructing the route that interested me initially most of all in these devices.
Unlike the iRoomba robots that flooded the market, the motion algorithm of which is based on indiscriminate throwing around the room at arbitrary angles (except for the latest models, where AI rudiments began to appear), iClebo and Neato robots are able to “build a map” premises ", as they say in the characteristics.
A wide-angle upward-looking camera is built into the upper part of the iClebo body, which allegedly takes pictures of the ceiling and thus determines the perimeter of the room. I conducted a test with a cloth rag on the camera, while the behavior of the robot did not change at all. As a result, it was felt that the camera is some kind of fiction, since the logic of constructing the trajectory of this robot, that with the camera, that without it, was in no way connected with the configuration and location of the rooms in the apartment.
Neato also has a completely different positioning system - a rotating laser range finder with an overview of 360 ° in the horizontal plane. The range finder is located in the "tower" of the robot, protruding above the upper plane. Due to this, the overall height of Neato is about 1 cm larger than that of the iClebo. Of course, I closed the Neato with a rag and a range finder, and the robot did become blind and moved quite uncertainly around the remaining sensors around the perimeter of the rooms, without starting to clean the middle.
The iClebo motion algorithm, though optimal, is rather primitive - parallel zigzags with obstacle avoidance. From the starting point, wherever it is located, the robot starts moving forward all the way into the obstacle, then turns around, shifting to the side by a width of one pass, and goes back again all the way. At the same time, he doesn’t look at the doorways, and if there are no obstacles on the next passage, he calmly leaves through the open doors to another room, and if there are several doors of different rooms on the same line, then continue until he reaches the wall or some item. All the actual intelligence of the robot lies in the fact that it remembers the cleaned and uncleared areas, returning to them afterwards. That is, if he went to another room on the next zigzag, he will remember that he has not yet found any obstacles to the left and right of his path, and will surely begin to “shade” the area, first on one side, then on the other, until this area it is limited around the perimeter by obstacles or areas already previously removed.
Neato also builds the optimal route, but the logic of its calculation is radically different. First, he travels to some arbitrary point of the room and scans the perimeter for a few seconds, slowly turning around its axis. Then, if you have not figured it out, you can drive away and scan again from another point. After that, the robot goes to the nearest wall and starts moving around the perimeter of the room. The direction of the bypass is always counterclockwise. The side proximity sensor is located only on the right side of the bumper. Having completed the perimeter bypass, the robot proceeds to zigzag cleaning of the middle of the room.
The Neato XV-11, which I had at the beginning, had an older firmware version 2.4. But in fact, this firmware was better. The main difference is in the quality of the definition of doorways. This robot always identified them correctly and cleaned each room separately and completely, and then moved to the next one. The only exceptions were small rooms such as a hall, bathroom or corridor. Neato considered them as one complex-shaped room, avoided them all around the perimeter, and then began cleaning the middle of each room in turn.
I got the newer Neato XV-21 with firmware 3.0, and here the algorithm has been greatly modified, now the robot often ignores doorways, but can easily divide the room into two parts and remove them one by one, and not even in a row. Unfortunately, the firmware update service only works up, and there is no possibility to downgrade. I could not find the firmware as such files, and the update process is performed only directly through the proprietary web service, which, after checking the current firmware of the robot, only reports the presence of an update, if any, and offers to update it. But the fact that, unlike iClebo, Neato’s main development is not hardware, but software, the update history of software has a dozen versions, and rumor has another update to version 3.1, which promises to be quite interesting.
iClebo always moves in parallel lines relative to its starting position, and never adjusts to the lines of the walls to move along them. That is, if you put the base at an angle to the walls in the apartment with simple rectangular rooms, then it will clean the entire apartment at an angle. Although, of course, the algorithm incorporates the behavior of “licking” an obstacle - when detected, the robot does not just turn around, but starts to go along the line of the obstacle until it moves to the width of its previous passage, and only then turns around and goes back. Thus, it will pass along all the walls anyway, but it can be “piecewise” and not in one pass.
Neato, on the contrary, always adapts precisely to the line of the walls in the room, regardless of its starting position. Having toured the perimeter of the room, the robot makes a decision on the direction of zagzags and begins to “shade” the area with parallel passages. Here the fundamental difference in the movement of these two robots is that iClebo at the end of each pass rests against an obstacle, and the Neato zigzags end earlier - without reaching the width of one pass of the robot, which it did when going around the perimeter. As a result, Neato is much less likely to dig into obstacles with a bumper.
Another reason that Neato rarely touches the walls and furniture with the bumper is, of course, in its omnidirectional laser rangefinder. This navigation device does not notice only obstacles located below its plane of view. And iClebo has only three point proximity sensors, one of them is looking forward and two to the side. For this reason, he often pokes bumper into narrow obstacles (furniture legs, edges of doors), if they are caught between the sensors. Of course, he leaves no dents and noticeable scratches, but he pounds quite loudly and perceptibly hits furniture.
With the search for the base, both robots usually have no problems, unless navigation has failed. Unfortunately, both robots suffer such a disaster. Here each model has its own characteristics.
Neato navigation failures are extremely rare, but possible. So far, few statistics have been collected by me for final conclusions, but there were already three different cases of failures. The first one happened literally out of the blue - the robot suddenly flustered, he randomly twitched left-right-forward-back on a small patch of completely free space near the wall. Then he stopped and asked to clear his way. There was nothing to clear, I just told him that I had cleared it, the robot again convulsed, but after a minute I was able to overcome my temporary ailment and left. However, the complications were not long in coming - the robot was completely disoriented and began cleaning the apartment again, not finding its base upon completion. The second case occurred after a robot struggled with an obstacle about a centimeter high in the form of a horizontal painted steel pipe (support for a clothes dryer). He could not see this low obstacle with his “eyes”, and the bumper with a mechanical collision sensor turned out to be higher than this obstacle. The robot drove onto the pipe in the front and began to slip the wheels - the weak “lugs” and the heavy weight of the robot did not allow it to be overcome. Realizing that something was wrong, the robot began to produce a variety of maneuvers, moving back, turning in different directions and again attempting to jump the pipe, then across, then diagonally, then in reverse. It happened for a long time, but without success. In the end, the robot spat on this occupation and went to get out, as it seemed to him, further. However, this long struggle turned for him again completely disorientation in space. The third case was completely incomprehensible - in the process of cleaning, I simply twice used the pause function (suspended the cleaning process), after which - again disorientation. Three cases in three weeks - a bit too much. Although the causes of the last two are removable. I hope for an improved firmware.
With iClebo, everything is much worse. So much so that, as a result, the first copy of this robot went with me for exchange, and the second copy - for returning to the store, as a product of inadequate quality. The reason for iClebo navigation failures is in a gradually digging error of parallelism of passes. The first copy of iClebo discovered this flaw on the second or third day. The problem progressed with each passing day, which ultimately led to the complete inability to use the robot. The error manifested itself at the time the robot turned to one of the sides - the turn was not always exactly 180 °, sometimes a little more. This led to the fact that the new passage was not parallel to the previous one, and so with each new turn more and more, as a result, the accumulated angle of error reached 45 ° and more. In the memory of the robot, this error was not taken into account, it still believed that it runs in parallel (why then the camera?), And as a result its map of passes no longer corresponds to reality. For this reason, the robot lost the possibility of further normal navigation, could not find unfinished parts of the apartment, which eventually remained uncleaned, and cleaned up other areas again, considering them to be a new area. In the worst case, the robot simply lost the possibility of further movement, moving to the mode of endless cyclic poking at the wall and yawing along it left-right in the area of ​​about a meter in search of the passage, which, in his opinion, must be here. This case usually ended with an error and a halt.
At the end of the 15-day period from the date of purchase, I returned this copy to the store, the marriage was recognized and a new copy was issued for replacement. The second iClebo showed itself significantly better in terms of navigation, and in the first days it behaved just fine. But, to my disappointment, after a few days, this specimen found the above-described deficiency, although much less pronounced. Most often, there were no problems, but sometimes there were also unmade pieces and walks along the walls in unsuccessful searches of the base.
This was the main reason for the abandonment of the iClebo Arte in favor of the Neato XV-21. Maybe I got on the defective batch of iClebo, as I didn’t come across reviews with similar problems. Therefore, the conclusion about the superiority of one navigation system over another in this case may not be objective. The store accepted this robot without any problems and offered to replace Neato for replacement, which I did.
Strictly speaking, iClebo Arte is not a vacuum cleaner, but a turboscale. The main cleaning effect is achieved by a well-made rotating brush with a correctly located scraper gum. Garbage is not sucked from the floor, and sweeps into the container. The suction power is so small that I never managed to find out where this air cleaner was blowing air, it was not felt by hands from any side, and constructively no outlet ducts or grilles were found. The “vacuum” effect, apparently, is used only so that the dust does not strongly scatter from the turbo brush and side broom-helicopters, of which there are two, and after lifting with a brush to the container did not fall back. However, the dust still scatters around. After cleaning, the bottom of the robot is completely covered with an even layer of the smallest dust. Also, its upper part is significantly gathering dust. When observing the robot along the floor against the background of something dark, it was clearly visible that it was moving in a cloud of dust lifted by brushes. But, nevertheless, in his container the vacuum cleaner after each cleaning brought a lot of garbage. Low suction power made this robot very quiet in cleaning. The noise from his work is comparable to a children's toy on batteries.
Neato opposite - it is a vacuum cleaner. It also has a similar turbo brush, but its suction power is incomparably higher - a powerful jet of air escapes from the back of the robot. Of course, it is much noisier at work, the volume of the TV has to be doubled. The previous model Neato XV-11 had only one brush in a set - with straight rows of plastic petals. This design created an additional unpleasant buzzing noise. The new model Neato XV-21 has already two brushes in the set, and the second with spiral rows of petals and bristles. This brush works completely silently, but the vacuum cleaner is still noisy. Unlike the iClebo, the Neato case is not dusty either above or below during the cleaning process. This vacuum cleaner absolutely does not raise clouds of dust around itself. However, the amount of garbage he collects is not much more than that of iClebo. Both vacuum cleaners clean coarse dust, wool, hair equally well, which is the bulk of the collected garbage. iClebo badly removes only fine dust.
With cleaning along the walls and in the corners of both robots problems. And the much more significant problems here are with Neato. This robot has no side brush-helicopters, and its main brush is longer than the body of the robot, 4 cm on each side. As a result, about 5 cm along all walls and furniture remain never cleaned. The suction power, although it is large, is not enough to suck debris from the side.
iClebo along the walls, dust rather does not remove, but scatters more side to side brushes. Something will be removed now, something tomorrow. But, at least, visible debris along the walls does not remain and does not accumulate. Unlike Neato, with which this garbage accumulates day by day, and there are already visible dusty streaks along the walls.
With corners at iClebo it does not matter, triangles with sides of about 5-7 cm remain inaccessible for cleaning. Neato has very bad corners. Its adaptive algorithm for avoiding obstacles and a dampish way of turning in the corners leave substantial uncleared pieces, both on the inner and on the outer corners. Hope again on the updated firmware, where, judging by the video, this problem is partially solved:
Neato also can not boast of a good cross, by this criterion iClebo also definitely wins. It weighs half as much - about 3 kg versus about 6 kg in Neato, its wheels are wider and have more effective “grouser”. I have never noticed that this robot somewhere stalled. Neato suffers from this misfortune all the time, she stalls when she tries to call on the rugs, on the high threshold. Often this ends with a rollback and the passage of such an “obstacle” diagonally. Sometimes this does not help, and then the robot stops and asks for a path to be cleared for it.
Both robots have some nice features that by and large should not affect the choice, but it is interesting to mention them from the point of view of usability. More of these buns from iClebo. First, it has a remote control from which you can perform all actions with a vacuum cleaner - set cleaning modes, start cleaning and pause, program a daily timer, and even just move the robot by pressing the direction and rotation buttons. It would seem a useless thing. However, it is extremely convenient in the case when the base of the robot is located in a hard-to-reach place, for example, under a sofa or under a table. In this case, you do not need to crawl behind the robot to clean its container, but you can simply call it from the console to a more convenient place. In addition, on the remote there is a button to return to the base command, through which the robot returns to recharge itself. In Neato, the team returning to the base is completely missing; he can return there only if he started from it and completed the cleaning completely and without problems.
Secondly, iClebo has a wet wipe mode. It is necessary to wet the microfiber that comes in the kit, put it on a mop, and mop on a vacuum cleaner. It is not quite convenient, and the overcoming obstacles mode in this case does not work, but there is enough for wiping one room - it collects just the smallest dust, which was scattered around and could not be sucked in. In addition, iClebo has a maximum cleaning mode, in which, after the main cleaning is completed along the optimal zigzag route, the robot repeats the cleaning along the classical chaotic diagonal route until the battery runs out of power. Thus, the quality of cleaning rises even higher.
Neato cannot boast of such a variety of functionality, it has only two distinctive features. The first is the ability to program the cleaning time for each day of the week separately, while the iClebo has only one cleaning time setting and will be cleaned daily at a specified time. The second is a graphical LCD display, in which, besides the menu, the robot displays a variety of messages and errors like “Please clear my path”, “I am returning to the base for recharging”, “Thanks for cleaning my container and filter” and the like. Trifle, but nice. In iClebo, the digital seven-segment display, errors are displayed in the form of digital codes that can be deciphered by the cheat sheet, located under the base cover.
Of course, the robot vacuum cleaner is not able to completely eliminate the manual cleaning of the apartment from life. There will always be places in which the robot will not be able to get because of its size. For example, a leg of furniture located in a corner of a room - for it, the robot simply does not drop in. But, nevertheless, now it is necessary to resort to manual cleaning much less often. If earlier we cleaned the apartment on average once a week, now it is enough to do it no more than once a month. For a month and a half of using robot vacuum cleaners, I had to resort to manual cleaning only once. For myself, I made an unequivocal conclusion that the robot vacuum cleaner is not just a curious toy, but also a very useful gadget in the house.
Interesting article about the internal structure of Neato, about the navigation algorithm and programming of the robot
For ease of comparison, a summary table of the main selection criteria:
A large mirror is installed directly on the floor at a right angle to the floor plane.
The cabinet mirror starts at a height of 84 mm from the floor.
The depth of the cabinet base is 40 mm relative to the mirror plane.
The height of the robot without a tower - 83 mm.
The height of the scanning plane of the rangefinder is 87 mm.
Here I will try not to focus on the characteristics, the review of structures and other facts that can be gleaned from numerous descriptions, official and not so much. I mainly want to share my observations on the behavior of two completely different robots, on the effectiveness of their work and ease of operation. That is, the information that I could not get in any other way than a personal test drive, and which I did not have enough to make a final decision. Perhaps someone this article will help determine the choice.
So, after studying the specifications, prices, the possibility of buying, reading reviews and watching the video, only two vacuum cleaners remained in my attention - the iClebo Arte produced by the South Korean company Yujin Robot and the Neato XV-21 from the North American company Neato Robotics ". The intelligence of both of these models is implemented on the basis of an embedded Linux computer. The first one in my house was iClebo Arte, I used it for about a month. Then I watched a rental Neato XV-11 for two weeks, after which I purchased the Neato XV-21, which I have been using for a week at the time of this writing.
iClebo Arte
Neato XV-21
')
The principle of navigation and the logic of the route
The main characteristic that allows, in my opinion, to recognize robots vacuum cleaners as just robots, is fully automatic intelligent navigation. And it was the logic of constructing the route that interested me initially most of all in these devices.
Unlike the iRoomba robots that flooded the market, the motion algorithm of which is based on indiscriminate throwing around the room at arbitrary angles (except for the latest models, where AI rudiments began to appear), iClebo and Neato robots are able to “build a map” premises ", as they say in the characteristics.
A wide-angle upward-looking camera is built into the upper part of the iClebo body, which allegedly takes pictures of the ceiling and thus determines the perimeter of the room. I conducted a test with a cloth rag on the camera, while the behavior of the robot did not change at all. As a result, it was felt that the camera is some kind of fiction, since the logic of constructing the trajectory of this robot, that with the camera, that without it, was in no way connected with the configuration and location of the rooms in the apartment.
Neato also has a completely different positioning system - a rotating laser range finder with an overview of 360 ° in the horizontal plane. The range finder is located in the "tower" of the robot, protruding above the upper plane. Due to this, the overall height of Neato is about 1 cm larger than that of the iClebo. Of course, I closed the Neato with a rag and a range finder, and the robot did become blind and moved quite uncertainly around the remaining sensors around the perimeter of the rooms, without starting to clean the middle.
The iClebo motion algorithm, though optimal, is rather primitive - parallel zigzags with obstacle avoidance. From the starting point, wherever it is located, the robot starts moving forward all the way into the obstacle, then turns around, shifting to the side by a width of one pass, and goes back again all the way. At the same time, he doesn’t look at the doorways, and if there are no obstacles on the next passage, he calmly leaves through the open doors to another room, and if there are several doors of different rooms on the same line, then continue until he reaches the wall or some item. All the actual intelligence of the robot lies in the fact that it remembers the cleaned and uncleared areas, returning to them afterwards. That is, if he went to another room on the next zigzag, he will remember that he has not yet found any obstacles to the left and right of his path, and will surely begin to “shade” the area, first on one side, then on the other, until this area it is limited around the perimeter by obstacles or areas already previously removed.
Neato also builds the optimal route, but the logic of its calculation is radically different. First, he travels to some arbitrary point of the room and scans the perimeter for a few seconds, slowly turning around its axis. Then, if you have not figured it out, you can drive away and scan again from another point. After that, the robot goes to the nearest wall and starts moving around the perimeter of the room. The direction of the bypass is always counterclockwise. The side proximity sensor is located only on the right side of the bumper. Having completed the perimeter bypass, the robot proceeds to zigzag cleaning of the middle of the room.
The Neato XV-11, which I had at the beginning, had an older firmware version 2.4. But in fact, this firmware was better. The main difference is in the quality of the definition of doorways. This robot always identified them correctly and cleaned each room separately and completely, and then moved to the next one. The only exceptions were small rooms such as a hall, bathroom or corridor. Neato considered them as one complex-shaped room, avoided them all around the perimeter, and then began cleaning the middle of each room in turn.
I got the newer Neato XV-21 with firmware 3.0, and here the algorithm has been greatly modified, now the robot often ignores doorways, but can easily divide the room into two parts and remove them one by one, and not even in a row. Unfortunately, the firmware update service only works up, and there is no possibility to downgrade. I could not find the firmware as such files, and the update process is performed only directly through the proprietary web service, which, after checking the current firmware of the robot, only reports the presence of an update, if any, and offers to update it. But the fact that, unlike iClebo, Neato’s main development is not hardware, but software, the update history of software has a dozen versions, and rumor has another update to version 3.1, which promises to be quite interesting.
iClebo always moves in parallel lines relative to its starting position, and never adjusts to the lines of the walls to move along them. That is, if you put the base at an angle to the walls in the apartment with simple rectangular rooms, then it will clean the entire apartment at an angle. Although, of course, the algorithm incorporates the behavior of “licking” an obstacle - when detected, the robot does not just turn around, but starts to go along the line of the obstacle until it moves to the width of its previous passage, and only then turns around and goes back. Thus, it will pass along all the walls anyway, but it can be “piecewise” and not in one pass.
Neato, on the contrary, always adapts precisely to the line of the walls in the room, regardless of its starting position. Having toured the perimeter of the room, the robot makes a decision on the direction of zagzags and begins to “shade” the area with parallel passages. Here the fundamental difference in the movement of these two robots is that iClebo at the end of each pass rests against an obstacle, and the Neato zigzags end earlier - without reaching the width of one pass of the robot, which it did when going around the perimeter. As a result, Neato is much less likely to dig into obstacles with a bumper.
Another reason that Neato rarely touches the walls and furniture with the bumper is, of course, in its omnidirectional laser rangefinder. This navigation device does not notice only obstacles located below its plane of view. And iClebo has only three point proximity sensors, one of them is looking forward and two to the side. For this reason, he often pokes bumper into narrow obstacles (furniture legs, edges of doors), if they are caught between the sensors. Of course, he leaves no dents and noticeable scratches, but he pounds quite loudly and perceptibly hits furniture.
With the search for the base, both robots usually have no problems, unless navigation has failed. Unfortunately, both robots suffer such a disaster. Here each model has its own characteristics.
Navigation fails
Neato navigation failures are extremely rare, but possible. So far, few statistics have been collected by me for final conclusions, but there were already three different cases of failures. The first one happened literally out of the blue - the robot suddenly flustered, he randomly twitched left-right-forward-back on a small patch of completely free space near the wall. Then he stopped and asked to clear his way. There was nothing to clear, I just told him that I had cleared it, the robot again convulsed, but after a minute I was able to overcome my temporary ailment and left. However, the complications were not long in coming - the robot was completely disoriented and began cleaning the apartment again, not finding its base upon completion. The second case occurred after a robot struggled with an obstacle about a centimeter high in the form of a horizontal painted steel pipe (support for a clothes dryer). He could not see this low obstacle with his “eyes”, and the bumper with a mechanical collision sensor turned out to be higher than this obstacle. The robot drove onto the pipe in the front and began to slip the wheels - the weak “lugs” and the heavy weight of the robot did not allow it to be overcome. Realizing that something was wrong, the robot began to produce a variety of maneuvers, moving back, turning in different directions and again attempting to jump the pipe, then across, then diagonally, then in reverse. It happened for a long time, but without success. In the end, the robot spat on this occupation and went to get out, as it seemed to him, further. However, this long struggle turned for him again completely disorientation in space. The third case was completely incomprehensible - in the process of cleaning, I simply twice used the pause function (suspended the cleaning process), after which - again disorientation. Three cases in three weeks - a bit too much. Although the causes of the last two are removable. I hope for an improved firmware.
With iClebo, everything is much worse. So much so that, as a result, the first copy of this robot went with me for exchange, and the second copy - for returning to the store, as a product of inadequate quality. The reason for iClebo navigation failures is in a gradually digging error of parallelism of passes. The first copy of iClebo discovered this flaw on the second or third day. The problem progressed with each passing day, which ultimately led to the complete inability to use the robot. The error manifested itself at the time the robot turned to one of the sides - the turn was not always exactly 180 °, sometimes a little more. This led to the fact that the new passage was not parallel to the previous one, and so with each new turn more and more, as a result, the accumulated angle of error reached 45 ° and more. In the memory of the robot, this error was not taken into account, it still believed that it runs in parallel (why then the camera?), And as a result its map of passes no longer corresponds to reality. For this reason, the robot lost the possibility of further normal navigation, could not find unfinished parts of the apartment, which eventually remained uncleaned, and cleaned up other areas again, considering them to be a new area. In the worst case, the robot simply lost the possibility of further movement, moving to the mode of endless cyclic poking at the wall and yawing along it left-right in the area of ​​about a meter in search of the passage, which, in his opinion, must be here. This case usually ended with an error and a halt.
At the end of the 15-day period from the date of purchase, I returned this copy to the store, the marriage was recognized and a new copy was issued for replacement. The second iClebo showed itself significantly better in terms of navigation, and in the first days it behaved just fine. But, to my disappointment, after a few days, this specimen found the above-described deficiency, although much less pronounced. Most often, there were no problems, but sometimes there were also unmade pieces and walks along the walls in unsuccessful searches of the base.
This was the main reason for the abandonment of the iClebo Arte in favor of the Neato XV-21. Maybe I got on the defective batch of iClebo, as I didn’t come across reviews with similar problems. Therefore, the conclusion about the superiority of one navigation system over another in this case may not be objective. The store accepted this robot without any problems and offered to replace Neato for replacement, which I did.
Quality cleaning
Strictly speaking, iClebo Arte is not a vacuum cleaner, but a turboscale. The main cleaning effect is achieved by a well-made rotating brush with a correctly located scraper gum. Garbage is not sucked from the floor, and sweeps into the container. The suction power is so small that I never managed to find out where this air cleaner was blowing air, it was not felt by hands from any side, and constructively no outlet ducts or grilles were found. The “vacuum” effect, apparently, is used only so that the dust does not strongly scatter from the turbo brush and side broom-helicopters, of which there are two, and after lifting with a brush to the container did not fall back. However, the dust still scatters around. After cleaning, the bottom of the robot is completely covered with an even layer of the smallest dust. Also, its upper part is significantly gathering dust. When observing the robot along the floor against the background of something dark, it was clearly visible that it was moving in a cloud of dust lifted by brushes. But, nevertheless, in his container the vacuum cleaner after each cleaning brought a lot of garbage. Low suction power made this robot very quiet in cleaning. The noise from his work is comparable to a children's toy on batteries.
Neato opposite - it is a vacuum cleaner. It also has a similar turbo brush, but its suction power is incomparably higher - a powerful jet of air escapes from the back of the robot. Of course, it is much noisier at work, the volume of the TV has to be doubled. The previous model Neato XV-11 had only one brush in a set - with straight rows of plastic petals. This design created an additional unpleasant buzzing noise. The new model Neato XV-21 has already two brushes in the set, and the second with spiral rows of petals and bristles. This brush works completely silently, but the vacuum cleaner is still noisy. Unlike the iClebo, the Neato case is not dusty either above or below during the cleaning process. This vacuum cleaner absolutely does not raise clouds of dust around itself. However, the amount of garbage he collects is not much more than that of iClebo. Both vacuum cleaners clean coarse dust, wool, hair equally well, which is the bulk of the collected garbage. iClebo badly removes only fine dust.
With cleaning along the walls and in the corners of both robots problems. And the much more significant problems here are with Neato. This robot has no side brush-helicopters, and its main brush is longer than the body of the robot, 4 cm on each side. As a result, about 5 cm along all walls and furniture remain never cleaned. The suction power, although it is large, is not enough to suck debris from the side.
iClebo along the walls, dust rather does not remove, but scatters more side to side brushes. Something will be removed now, something tomorrow. But, at least, visible debris along the walls does not remain and does not accumulate. Unlike Neato, with which this garbage accumulates day by day, and there are already visible dusty streaks along the walls.
With corners at iClebo it does not matter, triangles with sides of about 5-7 cm remain inaccessible for cleaning. Neato has very bad corners. Its adaptive algorithm for avoiding obstacles and a dampish way of turning in the corners leave substantial uncleared pieces, both on the inner and on the outer corners. Hope again on the updated firmware, where, judging by the video, this problem is partially solved:
Neato also can not boast of a good cross, by this criterion iClebo also definitely wins. It weighs half as much - about 3 kg versus about 6 kg in Neato, its wheels are wider and have more effective “grouser”. I have never noticed that this robot somewhere stalled. Neato suffers from this misfortune all the time, she stalls when she tries to call on the rugs, on the high threshold. Often this ends with a rollback and the passage of such an “obstacle” diagonally. Sometimes this does not help, and then the robot stops and asks for a path to be cleared for it.
Buns
Both robots have some nice features that by and large should not affect the choice, but it is interesting to mention them from the point of view of usability. More of these buns from iClebo. First, it has a remote control from which you can perform all actions with a vacuum cleaner - set cleaning modes, start cleaning and pause, program a daily timer, and even just move the robot by pressing the direction and rotation buttons. It would seem a useless thing. However, it is extremely convenient in the case when the base of the robot is located in a hard-to-reach place, for example, under a sofa or under a table. In this case, you do not need to crawl behind the robot to clean its container, but you can simply call it from the console to a more convenient place. In addition, on the remote there is a button to return to the base command, through which the robot returns to recharge itself. In Neato, the team returning to the base is completely missing; he can return there only if he started from it and completed the cleaning completely and without problems.
Secondly, iClebo has a wet wipe mode. It is necessary to wet the microfiber that comes in the kit, put it on a mop, and mop on a vacuum cleaner. It is not quite convenient, and the overcoming obstacles mode in this case does not work, but there is enough for wiping one room - it collects just the smallest dust, which was scattered around and could not be sucked in. In addition, iClebo has a maximum cleaning mode, in which, after the main cleaning is completed along the optimal zigzag route, the robot repeats the cleaning along the classical chaotic diagonal route until the battery runs out of power. Thus, the quality of cleaning rises even higher.
Neato cannot boast of such a variety of functionality, it has only two distinctive features. The first is the ability to program the cleaning time for each day of the week separately, while the iClebo has only one cleaning time setting and will be cleaned daily at a specified time. The second is a graphical LCD display, in which, besides the menu, the robot displays a variety of messages and errors like “Please clear my path”, “I am returning to the base for recharging”, “Thanks for cleaning my container and filter” and the like. Trifle, but nice. In iClebo, the digital seven-segment display, errors are displayed in the form of digital codes that can be deciphered by the cheat sheet, located under the base cover.
Summary
Of course, the robot vacuum cleaner is not able to completely eliminate the manual cleaning of the apartment from life. There will always be places in which the robot will not be able to get because of its size. For example, a leg of furniture located in a corner of a room - for it, the robot simply does not drop in. But, nevertheless, now it is necessary to resort to manual cleaning much less often. If earlier we cleaned the apartment on average once a week, now it is enough to do it no more than once a month. For a month and a half of using robot vacuum cleaners, I had to resort to manual cleaning only once. For myself, I made an unequivocal conclusion that the robot vacuum cleaner is not just a curious toy, but also a very useful gadget in the house.
Interesting article about the internal structure of Neato, about the navigation algorithm and programming of the robot
For ease of comparison, a summary table of the main selection criteria:
| Criterion | iClebo Arte | Neato XV-21 |
|---|---|---|
| Navigation quality | worse | it is better |
| Navigation failure rate | more often | less often |
| The frequency of collisions with obstacles | more | less |
| The quality of cleaning the middle of the premises | below | above |
| The quality of cleaning the corners and along the walls | above | below |
| Noise work | below | above |
| Cleaning time | approximately equal to | approximately equal to |
| Operating time from one charge | more | less |
| Patency | it is better | worse |
| Ability to update software | questionable | there is |
| Price | a little higher | just below |
Neato XV-21 test in a mirrored navigation
A large mirror is installed directly on the floor at a right angle to the floor plane.
The cabinet mirror starts at a height of 84 mm from the floor.
The depth of the cabinet base is 40 mm relative to the mirror plane.
The height of the robot without a tower - 83 mm.
The height of the scanning plane of the rangefinder is 87 mm.
Source: https://habr.com/ru/post/159965/
All Articles