News 2017: Samsung POWERbot VR7000

[RobotJunkie30]

Almost perfect. I have had LG homebot and vorwerk vr200. The difference is that I do not need magnetic boundaries anymore on high pile carpets. But there are issues with front bumper and obstacles. I have the high power model 130 w.

+ The wheels have the best design of all robots.
+ Navigation is ok in well lit areas.
+ I have seen run times of 2 hours on silent mode.
+ Nice looking and practical dustbin.
+ It attacks corners like a champ.
+ It can waccum high pile carpets where non of the other robots could.
+ Suction is great on 130w models.

- On spot cleaning mode it bangs into objects HARD. ( Programming issue)?
- Sometimes it stops perfect before and object like 1 mm other times it bangs into it.
- Still the waccum turns off if it is on slight angle.( for example waccuming around chair legs that are on an angle) this is really shame that it just dies not back of instead of just stopping.
- Scheduling on non wifi models in only once per. current day or every day. That is not good. It's a simple timer, comon Samsung.

[glnc222]

The way it is getting trapped on furniture legs is not clear. Perhaps this is the same problem addressed on the 9000 series in thread "tilt reduction mod". If the side tilts up, by performing a "wheelie" under power, the sensor for wheel extension can be tripped, as though picked up off the floor -- shutting down for safety. Holding the wheels in corrects this issue. It can take quite a number of runs before the precise interaction causing a problem is encountered.
Do you mean that slanted furniture legs are more likely to block the robot? They might not be able to engage the bumper from certain sides. Photo of the bot in the stuck position would assist analysis.

All the robots have some sort of problem in my experience, and quite often some simple addition can improve performance.

[RobotJunkie30]

The way it is getting trapped on furniture legs is not clear. Perhaps this is the same problem addressed on the 9000 series in thread "tilt reduction mod". If the side tilts up, by performing a "wheelie" under power, the sensor for wheel extension can be tripped, as though picked up off the floor -- shutting down for safety. Holding the wheels in corrects this issue. It can take quite a number of runs before the precise interaction causing a problem is encountered.
Do you mean that slanted furniture legs are more likely to block the robot? They might not be able to engage the bumper from certain sides. Photo of the bot in the stuck position would assist analysis.

All the robots have some sort of problem in my experience, and quite often some simple addition can improve performance.

Yes still slanted furniture causes problems. Out if 3 times it did it twice. It tilted a bit on the corner of the bumper and just stopped. It is like they only test the vaccumes in the lab. Not in real homes. It is a bit like wireless charging on my galaxy s6 the charger lit up the room in the dark and I had to disassemble it to kill the lights...Ill post fotos when it does it again.

[RobotJunkie30]

https://goo.gl/photos/PbF48Ct3CU5ER5sn7

[glnc222]

If you want to fix that see 9000 post http://www.robotreviews.com/chat/viewto ... 91#p138691
Not very difficult and removable. It does look a lot like the same situation and design fault (tilt with wheels extended by a "wheelie" under driver wheel torque, tripping wheel extension sensor for lifting off the floor).

When this is fixed, the wheels run briefly when picked up off the floor while stopped, and then will stop, instead of instantly stopping if running. Press start to clear display error msg., ready to carry. Never pick it up by the wheels which might turn (advised in their manual as well).

I think the side bumper redesign on 7000, with the brush forward, may not require the addition made on the 9000.

Since only one wheel is extended they might be able to fix in software, with more sophistication about those sensors, but they don't seem much into those old details. Were it not so easy to fix on the outside, I would look at the sensor within, but apparently not needed.

[edit] Note that even though the bumper may engage the chair leg and normally trigger a maneuver, the side lifting shuts down the bot before the maneuver occurs. Keeping the bot going either responds to the bumper or scrapes by under the opposite wheel traction or both. The wheels may also stall triggering maneuvers similar to bumper hits.
Only testing with the simple fix will tell how effective.

[glnc222]

Note robots, not just Samsung, are made to automatically return to the dock if disturbed in this position, so the wheels will turn when it is moved and maybe when picked up (and always when picked up with the modification). Shove off those pesky owners interfering with charging. Also assistance for the vision impaired.

[RobotJunkie30]

If you want to fix that see 9000 post http://www.robotreviews.com/chat/viewto ... 91#p138691
Not very difficult and removable. It does look a lot like the same situation and design fault (tilt with wheels extended by a "wheelie" under driver wheel torque, tripping wheel extension sensor for lifting off the floor).

Bumper70009000Untitled.jpg

When this is fixed, the wheels run briefly when picked up off the floor while stopped, and then will stop, instead of instantly stopping if running. Press start to clear display error msg., ready to carry. Never pick it up by the wheels which might turn (advised in their manual as well).

I think the side bumper redesign on 7000, with the brush forward, may not require the addition made on the 9000.

Since only one wheel is extended they might be able to fix in software, with more sophistication about those sensors, but they don't seem much into those old details. Were it not so easy to fix on the outside, I would look at the sensor within, but apparently not needed.

[edit] Note that even though the bumper may engage the chair leg and normally trigger a maneuver, the side lifting shuts down the bot before the maneuver occurs. Keeping the bot going either responds to the bumper or scrapes by under the opposite wheel traction or both. The wheels may also stall triggering maneuvers similar to bumper hits.
Only testing with the simple fix will tell how effective.

I think it is a easy fix with something glued on the sidebumper to increase distance from the bumper and objects. Last night it threw an error C05 something with bumper sensor so that cold be the reason it was banging the furniture so hard. Also last night it got lost and waccumed the same area 3 times as it was getting dark outside. I am thinking of going back to vorwerk as I never had any issues with navigation or robot shutting itself down. Its really a shame as the robot has better design better suction, ckeans edges better and can vaccum high pile carpets. Amazon were kind enough to send me a new one but I am holding on to my VR200.

[glnc222]

I think it is a easy fix with something glued on the side bumper to increase distance from the bumper and objects.

That is actually hard to do with the bumper so thin, and is unlikely to work. The wider bumper will just produce the same "wheelie" effect differently. I experimented a lot with the 9250 on this issue and made a number of measurements of the wheel extension sensor. Besides, a thickness addition to the bumper is likely to be broken off, unless screwed on. Lowering the bumper instead is possible with a complete wrap around plastic strip, strengthened by the bend around sides, but I did not find it helps with this simple error in the wheel sensors easily fixed (it is actually more important to lower the bumper on the Vorwerk and Neato, for climbing trap problems -- a commercial add-on even being sold by a German shop online). The Powerbot wheel sensor over-rides the bumper, which takes too long to move the bot in order to prevent the "wheelie" effect and tilt.

Shining a flashlight through the front smoke plastic might reveal whether the 7000 has the counter weights seen on the 9350.

Vorwerks are not available here in the U.S. but I gather are sold in Britain besides the Continent. The similar Botvac required enhancements to work on the carpet. If I have to fix one, it was easier to make the Powerbot work. It is all up to personal preference and circumstance.

The VR200 has a larger bin and is more matched by the larger 9000 Powerbot, were the full capacity needed.
I am curious what the sonar on the Vorwerk accomplishes compared to simpler Neato's, and there are no reports about it. The camera guidance on the Powerbot, along with other brands, has some requirements different from the lidar guidance, which in turn has its own limitations with reflective surfaces etc.

[RobotJunkie30]

I think it is a easy fix with something glued on the side bumper to increase distance from the bumper and objects.

That is actually hard to do with the bumper so thin, and is unlikely to work. The wider bumper will just produce the same "wheelie" effect differently. I experimented a lot with the 9250 on this issue and made a number of measurements of the wheel extension sensor. Besides, a thickness addition to the bumper is likely to be broken off, unless screwed on. Lowering the bumper instead is possible with a complete wrap around plastic strip, strengthened by the bend around sides, but I did not find it helps with this simple error in the wheel sensors easily fixed (it is actually more important to lower the bumper on the Vorwerk and Neato, for climbing trap problems -- a commercial add-on even being sold by a German shop online). The Powerbot wheel sensor over-rides the bumper, which takes too long to move the bot in order to prevent the "wheelie" effect and tilt.

Shining a flashlight through the front smoke plastic might reveal whether the 7000 has the counter weights seen on the 9350.

Vorwerks are not available here in the U.S. but I gather are sold in Britain besides the Continent. The similar Botvac required enhancements to work on the carpet. If I have to fix one, it was easier to make the Powerbot work. It is all up to personal preference and circumstance.

The VR200 has a larger bin and is more matched by the larger 9000 Powerbot, were the full capacity needed.
I am curious what the sonar on the Vorwerk accomplishes compared to simpler Neato's, and there are no reports about it. The camera guidance on the Powerbot, along with other brands, has some requirements different from the lidar guidance, which in turn has its own limitations with reflective surfaces etc.

The main problem are the table legs in our kitchen so I could cover them in something more slippery or something more "rubbery" to stop it from climbing, without modifying the robot itself.
I have never seen climbing or trap problems on the vorwerk I have owned VR200 for more than 2 years now. I did have a problem with our old bar stools with typical round "leg" and had to use magnets on it. I believe VR7000 has "tilt/angle sensor" and have seen it back off stuff before it got in trouble. Never had any problems with shiny objects on vorwerk. Maybe older neato products but by now it is solved I think. Navigation is definitely better than camera as it vacuums with impressive precision and logic.
Sensors on vorwerk are used to detect objects as it approaches it slows down and it rarely touches anything with front part of the bumper only sides. I had a botvac 85 and it "had to" touch everything. I will try with the new vr7000 when I get it from amazon as the bumper sensor apparently was not working properly C05 error. I really want to like it, and have my whole house vacuumed, including the high pile carpets. Navigation I can live with, and have it set to vacuum during day and the table legs climbing problem could be solved easily.

[glnc222]

Are you saying there was some carpet the Samsung can travel where the VR200 would not work? The VR7000 and VR200 have similar size wheels compared to the larger 9000 wheels. Some deep pile carpets defeat all the robots. Which 7000 model do you have, the full number?

[RobotJunkie30]

Are you saying there was some carpet the Samsung can travel where the VR200 would not work? The VR7000 and VR200 have similar size wheels compared to the larger 9000 wheels. Some deep pile carpets defeat all the robots. Which 7000 model do you have, the full number?

Yes actually we have 2 high pile carpets that VR7000 can vaccum that other bots could not do properly or got stuck. The main difference is in the way thread is formed on the wheels, the VR7000 has more traction with pointy "spikes". Samsungs has this nailed down now and it is way ahead if competition. The VR200 sidebrush would also tangle and break on thick carpets. This one VR2DM704IUUET.

[glnc222]

What country Samsung is this? Looks Belgian but they don't show a 704 among other 70x. Most countries are four digit 7040, 7070 etc. Samsung is messy with different country models.

Turn on some lamps to run at night. It does not take much light.

Limiting wheel extension (from 9000 thread)

[glnc222]

The different bottom of the VR7000 would require another way of taping cardboard into the wheel well. The cliff sensor is positioned inconveniently next to the wheel.

It may be possible to place a strip of tape on the outside of the case, going across the bottom and down into the well side at the front of the wheel, to hold in the cardboard against the wheel spring. Otherwise epoxy glue might be used (which removes cleanly from plastic) or the thumbtack fastening. [edit] A small right angle plastic bracket form cut from wall corner guard in hardware stores could help. Double stick tape could apply flat on the case bottom as well as the wheel well side, with the tape strip holding it down as well. Some sheet metal could also be bent into such a bracket the right size. Metal Mirror mounting brackets might be adapted, close to the needed size. Actually, a straight plastic strip, cut from packaging something, to extend over the lip of the wheel well to hold down the cardboard, under the tape strip, would help as well.

[RobotJunkie30]

Update

Definitely it bangs on anything that is black in front of it. Yesterday it just stopped and shut itself down in bathroom, I picked it up display came on. It did not bother to get back to charging station! Is this normal to search for it around the house as it just decides to stop random place?

[glnc222]

Of course it bumps anything black, not detected by IR sensors; the VR200 has sonar, as does the cheaper new Xiaomi.
When it stops on an error it may not return to the base. It shuts down when it gets trapped. Other robots may resume if released from traps, but not Samsung. The 9000's require a space at least 20 inches wide in which to turn, maybe 18 inches on the 7000, I would have to test. They are at least smart enough not to fuss forever in spaces too tight, and shut down after a while.
The VR200 might run on more carpets with the bottom slider addition used on Botvac.

[wazzuppi]

RV.0.0.54 released recently.

Not much changes to operating performance, still its a few things hard. But has been able to find the charging dock much easier.

[RobotJunkie30]

Another WTF moment. I think it would deplete it's battery doing this I found it after 5 min of this. I thought it cannot back up, but it can if you keep pressing the bumper so it is just a brogramming issue. I feel that the designers of this vacuum did a really great job but the programmers

https://youtu.be/HZ9VY_O1H5c

And the banging mode got enabled again (not black color objects)

https://youtu.be/MIwc0x0OveQ

[glnc222]

It is not just Samsung which uses obstacle detection only to slow down the robot for bumper contact, which is done to bring the intake as close as possible to the obstacle with bumper indication.
It does not always work and I add a thin strip of soft velcro across top of the front to protect the bot as well as furniture, especially those with recessed bottoms seen by sensors as farther away than the actual bump.

None of the robots can handle forests of straight chair legs perfectly; the round shapes, Roomba 980 may do it best. For best results the chairs need to be moved and ideally, leaned back on only two legs.
Straight leg chairs can also fail to provide the necessary clearance around the sides for the front intake, square bots, to turn, causing them to be trapped with the "standard substandard software".
Interesting this bot runs on the black floor which has been a problem for many (and cautioned in the Samsung manual); the polished surface, though, tends to reflect better than others.

[RobotJunkie30]

It is not just Samsung which uses obstacle detection only to slow down the robot for bumper contact, which is done to bring the intake as close as possible to the obstacle with bumper indication.
It does not always work and I add a thin strip of soft velcro across top of the front to protect the bot as well as furniture, especially those with recessed bottoms seen by sensors as farther away than the actual bump.

None of the robots can handle forests of straight chair legs perfectly; the round shapes, Roomba 980 may do it best. For best results the chairs need to be moved and ideally, leaned back on only two legs.
Straight leg chairs can also fail to provide the necessary clearance around the sides for the front intake, square bots, to turn, causing them to be trapped with the "standard substandard software".
Interesting this bot runs on the black floor which has been a problem for many (and cautioned in the Samsung manual); the polished surface, though, tends to reflect better than others.

Yes I will probably glue something on the bumper. Yes but after a restart it does not ram the chairs anymore so something is up! No problems even with black carpets and dark tile. It actually ramps up the suction on black caroets going from black tile. Definetly they upgraded the sensors.

[glnc222]

The IR proximity sensors measure the distance to targets, not just binary detection. So a typo or something in the software could mess up the usage.
It takes a lot of runs to encounter the "wheelie" problem of tilting with wheel extension sensor tripped (prior post above), and see whether fixed in software (hard to reproduce under manual control as sensors seem ignored then). That this is also software is shown by Neato robots flipping up and extending the wheels all the time without stopping. The company even mentioned integration of data from the cliff sensors and accelerometer for "off the floor" detection. When tying in the wheels on Neato, lifting off the floor is still detected, while the Samsung has to run wheels a bit before stopping.

I have noticed when Samsung is fumbling around under a cabinet turning several times before withdrawing, it at least gets out fairly soon compared to Neato's which could stay forever, constantly turning the wrong way and ignoring the way out. Yet the Samsung camera is probably disabled then while Neato's lidar is still effective. They could all do better if they processed the path into such spaces more thoroughly with higher level logic, and got out the way they went in. Such refinements do not seem to be their highest priority; adjusting color on the smartphone app sells more.

The Spring 2017 sale promotion discounts have ended end of May.