[Idea] Neato Self Emptying

[Scapegoat]

Hi guys,

Just a thought to throw out there that maybe I'll pursue if I have some free time. After seeing this old post: http://www.robotreviews.com/chat/viewto ... =4&t=19456 In which a forum member redesigned the Neato XV dirt bin, I started thinking about the BotVac Connect D-series dirt bins. I don't think they suffer from the same airflow issues that the Neato XV dirt bin had. However, one of the things that's a little annoying is emptying the dirt bin on a daily basis.

There are a few robots on the market right now, Roomba i7+, Shark IQ, etc which have a "Self Emptying" System (Video: https://www.youtube.com/watch?v=Et_ROh1XICY). Essentially when they're docked, the dock acts like a vacuum to suck the dirt out of the dirt bin into a larger bin.

I was thinking that since the top of the Neato's dirt bin is accessible (unlike other robots like the Roborock), it might be possible to design a new top for the dirt bin. One that has a hole in it and some sort of flap that seals (with a gasket) when the robot is running. Then modify/rebuild the dock to incorporate a cheap vacuum that turns on for a few minutes while the robot is docked which would suck the dirt out of the top.


I'm a bit new to Neato and I know this community has a lot of experience with them. So would love to hear your input.

[jdredd]

What is interesting is the VR300 has a port in the bin setup, to hook up to their VK200 vac.



Problem being in our case it seems, you need a way to make the bot think you took out the bin and put it back in.
As it lets off and then pushes the nub button.

Modification could be to tap the pins for that button.

2 contacts/wires as part of the new bin housing. Then your suction flap port whatever, mechanism to simulate removing the bin and putting it back in.

[glnc222]

Topic previously discussed in very old thread http://www.robotreviews.com/chat/viewto ... lit=+empty
One of the Chinese robots with this feature seemed to have simplified it by making the bin accessible from the bottom so the robot parking on the charger positions it over the intake for the stationary vacuum. They also used a hand held vacuum on the base which could be taken off and used separately for greater value.

[Scapegoat]

What is interesting is the VR300 has a port in the bin setup, to hook up to their VK200 vac.

Ah didn't know about the VR300, I'm curious how does that hole on the top manage to seal itself when the VK200 isn't in place? Is it some sort of door with a gasket on it?

Topic previously discussed in very old thread http://www.robotreviews.com/chat/viewto ... lit=+empty
One of the Chinese robots with this feature seemed to have simplified it by making the bin accessible from the bottom so the robot parking on the charger positions it over the intake for the stationary vacuum. They also used a hand held vacuum on the base which could be taken off and used separately for greater value.

Thanks for that link. Yes, it seems like the Roomba and the Shark IQ also have the robot drive over the intake. However, for easiest modification I think it might be best to do it from the top.


The electronics + programming portion would be pretty straightforward for me. The mechanical design is where I have zero skills.
Is there somewhere I can buy another dirt bin? I'd like to try hacking around. Designing and 3D printing a whole new dirt bin would probably be a pain (my CAD skills aren't great) and I'd rather not mess up my current one.

[glnc222]

Many Neato parts are on Ebay, including dirt bins for both series.

[edit] See later post on a pneumatic actuator eliminating motors
http://www.robotreviews.com/chat/viewto ... 36#p155336

Not sure I posted it but I did design a mod for the Neato consisting of a spring loaded cover for an opening in the top which could open just under suction pressure from an attached vacuum. An attachment would be fashioned, even from wood, maybe some tin can, for a canister hose vac with a larger opening than the Neato top so it would work with the robot varied in position on the dock; the bottom edge would be fitted with foam weather strip to make a seal against the top. So you would have to construct a sort of crane to raise and lower the intake on a hose onto the robot (and not disturb the charging connection). There are kits for making robotic arms which might be used. One way to make it is with a sort of cabinet which surrounds the dock open on the sides, like a bookcase, with a stick mounted to move up and down through a hole in the top surface, with the vacuum hose fastened -- straightforward woodworking. Small gear motors are available which can be controlled by relays or simple prototype circuits on perfboard. There are several robotics hobby stores with parts online. The canister vac can be controlled by a relay. Arduino microcontrollers and accessories might be used (for availability of accessories ready made, besides all the software support; otherwise the simple 8-pin microcontrollers probably suffice). Instead of a motor, a plunger solenoid might be used to raise and lower the intake. Some counter weighting may be involved. Lever action could also be set up. The robot docking must be detected, and the most reliable method I once used is to monitor the dock plate voltage; some experimentation required (see thread on Dock Based Clock). On my Samsung it is easier because the dock has a "docked" indicator LED which can be tapped.

[Scapegoat]

I did design a mod for the Neato consisting of a spring loaded cover for an opening in the top which could open just under suction pressure from an attached vacuum.

This sounds similar to what I had in mind, I'm curious where was the spring in relation to the cover? Was it there to push the cover down against the top of the Neato dirt bin?

[glnc222]

Sorry never did every detail. A hinged cover can be made with pivot pins on the side, or a small hardware hinge fastened in the center of side, cabinet hardware, some very small hinges, fastening with screws. A commercial fitting might use a thin spring made to be twisted, torsion spring, with wire ends fastened into material at the ends, on opposite sides of the hinge pivot pin. Tension needs to be low, so the door can open under top suction, and held down by Neato fan suction when cleaning. Spring wire can be found inside ball point pens with the push down/up feature. Lots of springs at distributors (Amazon, hardware depts.) are too strong, tension too high (made for doors etc., you need to see what is available). it would even be possible to use a rubber band fastened inside the bin at one end, the other on the bottom of the door near the hinge (one on each side of the door flap); this would allow wide latitude for adjusting tension. Maybe not as durable as a metal wire -- but cheap for experiments.

The edges of the door may have thin foam weatherstrip to seal when seated (that stuff can be trimmed with scissors). The seal could also be on the bin top with the door closing onto it. Just takes a lot of fiddling and imagination. Thin wood is available in square dowel in lumber depts. and thin boards, basswood or balsa wood in craft stores or online. I am in the habit of making things out of cardboard for experiments -- easily cut as needed (and use tape to make a hinge). Also scavenging plastic from packaging of various products (DVD/CD cases is a good source, besides those annoying clam shells sealing everything for store racks. Sheet plastic is available on Ebay. You can cut a neat hole in the top of the bin with a hole saw, and other ways of sawing. The cut out might make a door.

Whether an attached full size vacuum will empty the bin, with the air passages available in the robot structure, remains to be tested.

Another vacuum handling rig might be a vertical stand with a seesaw arm attached to raise and lower the vacuum hose; various leverages can be arranged. Plastic pipe in hardware stores, supplemented with tape, can make
fittings for vacuum hoses -- and posts for a stand.
Important the door seals well or the robot will lose suction for cleaning through it.

[edit] Any door fitting on top must not protrude vertically very far or it will interfere with the lidar beam.

[glnc222]

Addendum: rubber band -- I think this would work best were there one in the center running from some point along the front-to-back center line of the door, to a point back behind the hinge inside the bin. Perfect for adjusting tension, and bends around the hinge neatly when door opens. So a stretching metal spring could do the same but hard to get the right tension. For experiments the rubber band can just be secured with duct tape (Gorrilla type best).

[glnc222]

Rubber band might best be fastened with a staple into the door material, thin wood or cardboard, if you have the right kind of stapler than unfolds for tacking. There are harder grade staples than regular, actually necessary for use in the fancy power-ease snap action easy press type. In a plastic door tiny screws could be used, available in better hardware stores, no. 2 etc.

[glnc222]

Canister vac might be controlled by a small lever action microswitch on the end of the hose activated pressing against the Neato top when lowering the hose, instead of extra electronics for computer control. Then the only thing needed is raising and lowering the hose. To do this, a motor might be fastened to a post turning a spindle winding and unwinding a cord pulling on a lever arm on the post. Or a pulley atop the post and on the motor, with the arm attached to one side of the loop. Counter weights could make the required torque very low, like doors into secret passages in the castle.

[glnc222]

A 4 1/4 in. wide funnel at amazon might be used for the canister vac attachment to mate with the Neato, cut off to fit the vacuum hose. The tab on the side could help mount a microswitch to control the canister vac.
Larger funnels are also listed.
https://www.amazon.com/Winco-PF-8-Plast ... 539&sr=8-9

For raising and lowering the vacuum hose or pipe, robotics supply includes linear actuators, e.g. at servocity.com
https://www.servocity.com/motors-actuat ... -actuators
Might work with the cabinet and plunger design.

[Scapegoat]

For raising and lowering the vacuum hose or pipe, robotics supply includes linear actuators, e.g. at servocity.com
https://www.servocity.com/motors-actuat ... -actuators
Might work with the cabinet and plunger design.

Linear actuators are generally quite expensive. There should be another way to raise/lower the vac pipe maybe with a servo or a motor and pulley system. Or maybe it could have a spring (low tension) such that its state with 0 tension is expanded and the servo or motor can pull up on it when vac is undocked?

[glnc222]

I notice that with a shelf above the robot, plastic pipe (sold by the foot) can go through a circular hole with a funnel on the end fixed up with foam weatherstrip, to make the vacuum connection; maybe an extra bit of board on top to thicken, make a longer hole mounting for stability. The microswitch to control the canister vac can be tapped with resistors and maybe diodes to provide a low voltage, logic level contact sensing for the microcontroller, to tell when the pipe is bottomed onto the robot. Another microswitch at the top end of the pipe under the shelf tells when the pipe is raised (what's called limit switches).

So you could mount a pulley above the shelf on a stanchion or backboard above the shelf, and on the motor atop the shelf, with the belt (rubber o-ring) attached to the pipe to move it up and down, DIY linear actuator. Or a cord could be wound on a spindle pulling it up, weighted to pull down. See what supplies are available at servocity.com and similar hobby stores. A gear motor is likely wanted for low rpm and strong torque, compared to say the side brush or lidar spin motor (CD player motors).
Precut lumber in Home Depot and Lowes. A piece of 1x2 stick could be fastened by screw from underneath the shelf for a stanchion. Maybe two layers glued together, or a chunk of 2x4 for thicker base.

[glnc222]

All sorts of Arduino specific stepper motors are listed on Ebay. I am not familiar with all the Arduino stuff, just used one once with a pressure sensor for analyzing robot vacuum suction. Stepper motors are controlled differently than continuously turning motors, typically used for all the mechanisms in printers for example. There must be tutorials especially with Arduino. Very cheap Arduino's are on ebay from China compared to U.S. dealers.
A popular U.S. store is SparkFun.com.

[glnc222]

Instead of dropping under weight the pipe can be pulled down by a rubber band, and lifted by the motor. Winding a cord eliminates pulleys.

[glnc222]

Slice of a thick round dowel (up to 1.5in in lumber dept.) might be fashioned into a winch drum drilling a hole to fit a stepper motor shaft, maybe with a set screw through the side. In the hole for the pipe in the shelf, file a narrow slot on the inside of the hole to pass a lifting cord through to the top side where the winch is placed, so the lift pulls straight up. Where the vacuum hose attaches to the top of the pipe, some plumbing right angle fittings might make a U shape to attach the hose pointing upwards. Devil's in the details, whatever bracing is needed.
A hole saw can also cut out a thick round piece from a board instead of slicing a dowel.
The pipe probably has to move up and down about 3 in. minimum, to clear the top of the charging dock so the patterned reflective front behind smoke plastic is visible to the lidar.

[jdredd]

My ideas..





This was always my idea... at least when it comes to the canister suction.

Problem would be gravity. Would need almost shop vac power to get the suction/seal required for
this kind of design to work.

1.25", 1.5" , 2.5" is most sizes I think... so a 1.25 hole in canister... then 2.5" suction tube.

Would need a way to insure alignment between botvac and suction tube device so it lines up with a little bit of
wiggle room with 125" hole and 2.5" suction.

I would then gut a charge base station to make a whole new one and use the charge plates as the triggers to do the whole thing.

If wanted to get real fancy, use a Pixy vision sensor and some alignment dots of sorts on top of canister to signify the alignment when docked. If crafty give the lowering/suction tube a few degrees of tilt/pivot action.

Or worse case.. wait for Neato to play catch up and make a new model with this already out of the box...


As for the push button for bin being opened closed... just make a small circuit to simulate the button push.
when the 2 wires for charge base are powered, kick off a timer that is soldered to button to connect the points for a moment and off. That way it thinks you removed and put back the bin.

[glnc222]

As to the bin access hole and cover, what I would try is fitting Gorilla duct tape around the inside protruding 1/4 in. inside the hole edge, and covering that part with foam weatherstrip onto which the cover plug would rest. If something stiffer is needed I would cut out a hole from thin plastic and tape or glue that on the underside of the hole to the same effect, finishing the inside edge of bin hole with foam. Maybe just foam strip, left a bit thick, could be stuck around the hole protruding inside 1/4 in. and paint over the exposed glue part within the hole, or put some thin foam on top of that. Keep the cover flush with the bin top as much as possible for lidar clearance. You can measure the clearance available with the lidar scan in the Neato Control program.

[glnc222]

Limit switches: the upper limit microswitch is mounted under the shelf in fixed position for neater wiring. A bit of wood or something is screwed onto the pipe to bump the switch, or maybe a strip-and-screw circular hose clamp, whose position can be adjusted on the pipe. Lots of microswitches on ebay besides distributors like DigiKey.

Another thing which might be used for the vacuum fitting to the bot is the rubber bottom of a toilet plunger; mine has a 5.5 in. diameter, unless too big.

[glnc222]

CORRECTION: the bottom limit switch cannot directly control the canister vac because the suction will "glue" the pipe down on the bot preventing a small motor from lifting it off. An Arduino appliance control relay will be needed, plenty available. Vacuum must be turned off before lifting the pipe -- time delay while vacuum slows.

The charging base must be fastened in place on a cross beam or backing of the shelf/cabinet, for stability under the vacuum fitting contact pressure. Possibly screwed to the beam, or with 3M Command picture hanging strips, similar to velcro but stronger, could be removable when wanted for other uses. Always paint wood before attaching Command strips, for a proper gluing surface. Another method is to fashion keyhole slots in the back of the dock, holes passing the head of a screw then slid down into a holding slot (made with a smaller drill bit or file) -- then just pop on and off, vs. tedious prying apart locking strips.