Page 1 of 1

Help with some robot research

PostPosted: September 29th, 2017, 3:19 pm
by robojoho
Hello - I'm a robot vacuum newb and I'm currently working on a project for a robotic vacuum. I'm looking to learn about peoples' experiences with their robots and would like to gain insight from the community.

If you're open sharing your experience, I'd love to connect - please shoot me a PM or just drop me a note on this thread. Also, no worries if you aren't a current robot vacuum owner. I would love to hear from both robot vacuum owners and non-owners!

Thank you in advance - hope to talk to lots of you soon! :cheers:

Re: Help with some robot research

PostPosted: September 29th, 2017, 7:42 pm
by glnc222
My experience with robots is pretty well documented in the Neato and Samsung forums over years, and a review section comparing Roomba and Neato. The products have advanced slowly over a long time since the first Roomba's nearly twenty years ago.

The essential point has been that these products have peculiar gaps in their design preventing them from working with a variety of furniture and flooring features, but oddly can be improved in some cases with user additions or modifications (which can be an entertaining hobby). Some were added by makers in later models. This raises a question why such features were never considered by the suppliers in the first place, what sort of business this is.

Another major point is these low power devices are mostly not equipped to clean embedded dirt in carpets, for which the much more powerful standard vacuum cleaner was invented a century ago. They work well as hard floor sweepers, with such flooring more popular nowadays. Some have recently reached the point of supplanting the standard vac and comprising a substantial share of new vacuum sales -- probably explaining a raft of new entrants.

A different slant is experience with the robot makers compared to with their robot products, difficulties in delivery, maintenance, and support.

Re: Help with some robot research

PostPosted: October 10th, 2017, 8:15 pm
by robojoho
glnc222 wrote:The essential point has been that these products have peculiar gaps in their design preventing them from working with a variety of furniture and flooring features, but oddly can be improved in some cases with user additions or modifications (which can be an entertaining hobby). Some were added by makers in later models. This raises a question why such features were never considered by the suppliers in the first place, what sort of business this is.

In your opinion, what are some of the more significant modifications done by makers to later models? Interested in the bigger pain points that weren't addressed by suppliers, from your point of view.

I'd like to dig in on the customer experience slant you mentioned around delivery, maintenance, and support. Is there anyone you feel is doing it right? I'm also curious to learn more about the level of interaction or relationship you would want to have (if any) with the makers of the product beyond standard customer service and maintenance, and whether it would impact your decision in selecting a device.

P.S. - I've checked out some of your reviews and posts, and they've been very helpful and insightful. It will take some time to get through all of the posts, of course. ;)

Re: Help with some robot research

PostPosted: October 11th, 2017, 1:59 am
by glnc222
If appropriate it would be interesting to hear what sort of project you are working on.

Very few customers modify their robots, as it is sort of a hobby among skilled craftsmen, electronic engineers and technicians, often retired. This is not a general consumer website, more a technician and hobbyist site (though as a result we have the most information...).

Atop the Neato forum here you will see an index thread i maintain which tries to catalog all the modifications threads on the Neato, many which I did myself, usually with no followers.

Modifications were originally due to very poor, cheap construction of the robots, and absence of useful features slowly added to new models over a decade (they started in 2000 with Roomba). So member Vic7767 here had a business selling ball bearings to improve construction of Roomba's flimsy product prone to failures.

One mod of interest was due to earlier models using rapidly worn out NiMh batteries instead of lithium batteries only common last couple years; these bots make heavy duty battery use compared to other devices, more like an electric bike, or a power tool, and needing almost annual replacements. So early mods included ways to assemble your own longer life lithium battery and finesse the electronics for charging it involved, still used by some on old Neato and Roomba bots. One company began selling replacement lithium batteries for the old bots with included adapter electronics, and then all the new model bots came with lithium. Vorwerk was the first with its earliest Neato upscale version. Lithium became practical with advances in the chemistry making them safer than the kind in laptops and phones, for their large size more like car batteries, with power tools going to it a few years ago. LiNMC chemistry.

Early robots did not have WiFi communication and a number of electronics hobbyists and engineers worked on adding components to Neato robots for this purpose, one of which is now a permanent thread atop this forum. Lots of old models still in use. Engineering students can also have an interest in such projects. Now with all the WiFi equipped models, programmers, hackers are attracted to fiddling with the app software (Xiaomi originally had not English language and all sorts of hacks were provided by software expert users). The Alexa integration now popular is supported by Amazon for programmers, DIY, so there is a potential for a lot of hacking there not yet seen.

Two Neato mods have been used by several users. First is adding a slider plate underneath to improve movement on some carpets on which the robot could not move (robots have never been able to move on deep pile carpet or shag carpet much; some carpets are also slippery). Today you see the recent Xiaomi Neato competitor has a floating intake design preventing it moving on lots of carpet even regular Neato's navigate. But I like the big Samsung robot with larger wheels which handles carpets better without modification.

The second Neato mod, and for other brands, used by others has been adding a strip to the bumper to lower the bottom edge, so the robot does not climb onto low things on which they get stuck like a cat in a tree. Especially cantilever and tubular leg chairs, like from Ikea. Finally a German enterprise started selling a ready made self-stick strip for Neato's and Roomba's. The designers seem oblivious to this need.

The early Neato's lacked a side brush so I added one, but only a couple other people have tried it. They were added to the Botvac second generation several years ago. This allowed discover such side brushes are useless on carpet, and have other problems. Popular for hard floors though, customers want their corners cleaned. Samsung has made a different solution in their new 7000 series, with a suction direction shutter on the front.

Design defects continue even in the Samsung I use. I have widely publicized a software defect which causes the Powerbot to shut down in certain contacts with furniture legs, easily worked around by disabling a sensor detecting when the wheels are extended as when the robot climbs, in the suspension (just tape it in or whatever, "Side Bumper Extension and Tilt Reduction Mod" Samsung forum. There is also a tilt reduction mod for Neato's. Another Samsung defect is insufficient bumper length along the side, allowing mine to climb onto floor lamp bases and get stuck. The Guild of Modifiers to the rescue. No takers anywhere else have reported though. Improved in the newer 7000 series started this year. The way the companies work, I will be surprised if Samsung ever fixes the software error, though I finally got it forwarded to headquarters via their own forum -- if it goes anywhere in that huge company bureaucracy. They put their resources into new dubious things like smartphone apps they think will attract customers -- whoever has the fanciest app wins...

This is why there is interest in the work of member third_deg here, former Roomba designer starting his own company, Third Degree, if it ever happens; maybe he will get it right for a change. Slow process filing patents.

Another Neato mod only a few have used is for the robot getting stuck by pressing itself underneath a sofa or something, an over head pinch trap. Someone market actaully a little stick on thing for Roomba's with the same problem to prevent this, and I showed how to easily add a wire loop to the Neato bumper to prevent this. Then Vorwerk added an overhead sensor onto the laser turret like they should have -- and Neato's models still lack it. The camera guided bots are less affected because of better optical sensors. An old expensive Karcher self-empty robot had top sensors for it. Some of these problems are addressed by raising the problem furniture a little with wooden fittings etc. instead of fixing the robot.

Now your second question about customer relations and what not, I am not sure entirely what you mean. First notice that most of these robots, lacking washable filters like Samsung, Dyson and some, require buying regular supplies to make them work -- not to mention previously, annual expensive battery packs. So there is an issue with whether a company can be counted upon to supply these things for several years use of an expensive appliance.
Beyond these "consumables" though is an issue of spare parts to repair broken machines, electronic circuit boards and what not. Samsung does a full scale appliance standard parts supply. I am not sure about Roomba, but Neato Robotics supplies parts only to warranty service contractors, good only for year. Lots of old Neato parts get sold on ebay, used, but not for newer models. Of course, technical skills are needed to make repairs. Yet there are small businesses (member Medtech here in Russia repairs, as does Vic7767, and Shirgal in Israel) -- where do they get parts?

There are often questions about technical issues with the robots, being both complex and limited in use various ways not disclosed, with carpets and so on, and some makers, like Samsung, have customer advice forums on their websites. Get the community to respond besides bothering customer service reps, common throughout the electronic tech industry now.

What about warranty repairs? Neato and Vorwerk are reported slipping badly with dubious outside contractors.
What if it is out of warranty (post your problem here...)?

For example, a question raised a few times was how can you schedule the robot to make two cleaning passes through the house instead of the one supported in the software -- if customers even realize that limitation. A trick was found for this on Neato's.
Lots of questions here are "this or that doesn't work right, what can I do?" as if they did not bother to ask customer service. Well, customer service reps are not supported well by the engineers, and often know nothing more than the user guide info. A common problem in the whole very complex tech device industry, and squarely management's fault, not the tech rep's.
I use a prepaid Tracfone cell phone because not calling a lot so cheaper, and if you look at their site they have all sorts of tutorials on usage of the different smart phones, which are very complicated -- virtually PC's... I don't see that at vacuum robot makers.
Of course, if they made better robots with fewer limitations on use their would be fewer questions. The companies don't want to advertise their limitations, it is difficult to do so without denegrating the product. The software industry has long had a different attitude, with the frequent updates to fix identified problems etc. -- all due to business, commercial use with maintenance operations, the complexity of those systems etc.
It has always been a matter of cost -- a good job could be done with a $10,000 industrial class robot, but the issue is what can be done at a consumer device price? Not everything it turns out, leading to all sorts of things like this website.

Re: Help with some robot research

PostPosted: October 12th, 2017, 3:23 am
by glnc222
One area I forgot is the matter of software improvements over time, updating customer's robots. This is now done over WiFi on the newer models. However, these updates appear to be limited to technical bugs in the complex WiFi communications itself, the newest software addition. Yet the cleaning operations of the robots are complex and subject to improvement. For example, in one model Samsung introduced a procedure to poke three times into corners for better cleaning. Samsung did not, however, add this behavior to previous models in any update. The practice of companies is to sell new software only through new hardware. None offer software alone for a fee, to improve old models. You have to buy new hardware. This is not uncommon with small digital electronics accessories, but with an expensive appliance there are questions. When Neato Robotics was a start up, it followed the business model used in software one could call work in progress, where the software was improved over time and updates were supplied to customers. Ongoing sales fund the work of improvement, and the updates are a marketing appeal. The early robots could be updated with USB cables. After several years though, Neato discontinued supplying updates, except over WiFi, and I have not heard of any special new features offered that way. Quite a bit of hacking, by Russian programmers, was done to make available old updates for older models still in use which were never updated by owners.
Some of Dyson's advertisement videos mention future improvements to be supplied but they apparently refer to features of their smartphone app. I once asked them whether they intended to update the internal firmware of the robot and got no reply.
I have previously mentioned a known software fault in the Samsung robots, rather simple. If it is ever fixed it might be expected only in new model robots, instead of fixing those already owned by existing customers.
So a certain skepticism is in order when dealing with any of the products in this field. My business experience, dealing with many industries as an economist and banker, makes me doubt the quality of the marketing strategies used by these robot companies. They do not seem very devoted to satisfying the customers in all the ways possible and practical. They appear to occupy ephemeral positions in marketing and branding, not expecting to be in business over a long term. I suppose unless they see some major shift in sales volume the executives are not interested.
Still there are some attractive products, such as they are at the moment.

Roomba 595 Ball Bearings?

PostPosted: November 12th, 2017, 2:42 pm
by leonicholson
Do you have any experience with ball bearing replacements? I see them advertised on Ebay as sized for no hassle replacement.

Is it advantageous to switch out the bearings?

I currently have two of the the little buggers (595's) running every night. They coexist happily and pick up all the dog hair, etc. :)

Re: Help with some robot research

PostPosted: November 12th, 2017, 3:42 pm
by vic7767
Those bearings fit one specific Brush deck (Black) color in the Roomba CHM. Roomba brush decks are manufactured in Red, Green, Black and Grey. What color is your 595 Brush Deck?

Re: Help with some robot research

PostPosted: November 12th, 2017, 5:06 pm
by leonicholson

That wouldn't work then. Mine looks like this.

Re: Help with some robot research

PostPosted: November 12th, 2017, 7:59 pm
by vic7767
There is a ball bearing mod for your 700 CHM, you can read about it here: