RC and Autonomous Pool Skimmer

[behrendfry]

I am a junior EE and I have been tasked to build a pool skimmer based on a clients patent. The skimmer looks like a shark and the mouth is open with a net inside to catch debris. The shark moves via a water pump configuration. A water pump pulls watr in through the mouth and pumps it out the back hrough a tube that is connected to a stepper/servo motor to control direction of the device. There is a valve that will close to divert the water towards the front for reverse. My task s to make this shark pool skimmer autonomous. I was thinking of putting 3-4 sensors on the head of the shark to detect objects in front of it. Using a micro-controller I would have the shark turn left/right/or reverse depending on what sensors were triggered. I was looking at the parallax Basic Stamp Microcontroller. Does anyone have suggestions or thoughts? Possible sample code to play with. I would have to conrol 1 stepper/servo motor (steering) and 1 valve for reverse operation and possibly a valve to control the water flow for speed variations. Any suggestions or helpful sample code and schematics would help.

 

[The Mad Professor]

Interesting idea , conventional wisdom in the field of pool cleaning is to move the water and with it the floating debris to a static net/filter.

However I can see the appeal of a robotic leaf hunting shark..

Why not just make it reverse and turn when it bumps the edge of the pool?

 

[mramos1]

Water and electronics.. Hope this is theory and you are not going to make one (silicone glue)..

The leaf hunting, I do not have a pool, but some float and some sink. So something that runs on the top and bottom would do a good job.

Is your unit going to skim the top only?

 

[behrendfry]

Yes, this unit skims the top of the pool.

 

[akg]

the project sounds interesting , what about having a look at "roomba's " concept. u'll only need 'bump' sensors.

 

[mramos1]

IR outside might be a problem, but a better option if in a pool. Ultrasonic is an option, like the backup sensors, but they would need to be sealed?

Pulsed laser would look cool in the eyes, but then all the warning labels and power required.

 

[JRoque]

Hello. Sounds interesting. I would simply make it swim in wide circles so when it hits a wall, it will slowly turn away from it and continue on. This would save you object sensors and reversing valves. By the way, if the shark's mouth is open when you reverse, won't the leaves get out?

Most, if no all pools have a skimmer built-in so practical market might be limited. I know some folks will buy to see the thing swim around though. I bought one of those automated cleaners a couple of years ago - an Aquabot model. It was the worst $1500 I've ever spent. It did not cleaned the pool like they claimed and it broke after 4 months when used just once a week.

Instead of a shark, I would buy a catfish. A device that goes underwater and brushes the surface. Not just vacuum it, but brush it. Even when the Aquabot did it's best, I always had to brush by hand after it. If you come up with such device, sign me up for one.

JR

 

[dknguyen]

I don't know why people don't like sonar. In air, yeah sure use radar or lasers if you can afford it. But if there's a liquid medium around...USE SONAR! Liquid sonar is way more reliable than air sonar and is more reliable than IR and lasers, especially if the wave has to travel through the water. Why on earth would you use it as your backup? THere's a reason submarines don't use radar or lasers.

YOu can also use the liquid sonar to detect floating leaves...you'd have a hell of a time doing that with EM waves because you the wave might leave the water and hit the pool edge or the water might be too murky (or if you are submerged, the electromagnetic wave just simply won't travel in water as far as sonar.

(Yes, they need to be sealed and specifically designed for water us- the underwater type) You can get them pretty cheaply if you know where to look.)

Item 2210: https://www.hobbyengineering.com/SectionS.html

 

[The Mad Professor]

Refresh my aging old brain please , how does one distinguish a two dimensional object such as a leaf floating on the surface of a fluid medium using sonar?

 

[dknguyen]

I figure that the pressure waves will echo off of a floating leaf sitting in the water and you should be able to detect it if the receiver is sensitive enough. I guess it would have to be pretty sensitive.

 

[The Mad Professor]

I figure that the pressure waves will echo off of a floating leaf sitting in the water and you should be able to detect it if the receiver is sensitive enough. I guess it would have to be pretty sensitive.

Indeed it would , for what do estimate the difference hieght to be between the leaf and the air/water interface?
Less than a millimeter shall we say , on a surface that is far from being stationary, it's not imposibble but it is a big ask to implement using hobby sonar designed for simple range finding. The return echo charecteristics may be used to distinguish between leaf and the water/air interface but again this adds another a level of complexity that may push the project overbudget.

 

[philba]

For surface debris, the reflection off of ripples will mask everything else out. the pool will never be calm with the shark cruising around. I think random direction will be as effective as anything else.

I own a pool and my skimmer is built in. a pump circulates the water and causes most debris to end up in the skimmer basket. There are corners where some debris gets "stuck" (eddies). If the pool gets used a lot, the debris gets pushed out of the corners but if it sits unused for a few days, the corner debris builds up. I'd design a mobile skimmer to find corners and thus clean them out.

frankly, a more usefull pool-bot would be one that cleans the bottom of the pool. there is always some junk there that needs cleaning.

 

[behrendfry]

Update

Thanks for the comments and ideas. Having the shark swim in circles could be a possibility but the issue with that is the ladder going into the pool, the shark could get stuck in the space between the ladder and the pool side. This of course depends how nice of a pool you have. So some sort of sensory is needed to control the device.

The idea of my client is not to only have a shark body but make it removable to attach different "skins" like a catfish or an alligator.

I thought about the SONAR idea but the cost is already an issue so any additional cost that may not provide a huge benefit probably will not be accepted into his design.

The design is just to skim the top of the pool and the main purpose is to skim and the added attraction to have a shark floating around the pool is what the client is marketing for. I will keep the post updated with the progress of the pool skimmer. Hopefully weekly updates depending on my time.