I have been working on a sensor for a while now that will detect the % of fill-rate of any bin it is put onto, using an ultrasonic sensor. One major struggle I've been dealing with is that sometimes, due to the amount of filth, dirt and grime inside a bin, the ultrasonic sensor will get contaminated with dirt and cannot take any valid measurements anymore.
Makes sense of course logically, and it's an issue that's present with any sensor brand. The one I use right now is the MaxBotix TrashSonar MB-7137.
I was just hoping there's some kind of loophole to this I'm missing still. Could an ultrasonic sensor possibly be configured so that it would ignore any returning echo that would come from an object too closeby? Are there any other range-finding sensors that would somehow be more resilient with this?
Alternatively, it'd mean I would have to go out every now and then to clean off the sensor in person, which is fine, but I just want to see if there is a better solution out there still.
INTP:
I wonder what kind of garbage is so effusive that it slathers up a sensor from underneath.
Any kind that contains moisture and attracts molds will do the job. Such as kitchen or garden waste.
Such garbage will also create a lot of water vapour which will condense on the cooler lid (and the sensor).
The inside of a garbage bin is not a very nice or forgiving environment!
groundFungus:
Could you protect them with a movable cover? Have a small servo to open the cover long enough to get a range and then close it.
Unfortunately, that's a direction I do not really want to head for, because:
(1) It's currently an extremely energy-efficient device that can run off of a single D-cell for 4-5 years. A servo would drastically reduce this lifespan. I also plan to get more of these devices made and the logistics of battery replacements would become tiresome (compared to a quick wipe of the sensor surface)
(2) It's meant to be IP67-rated, and it will become really tricky designing an IP67-product with moving parts at an affordable cost.
I would probably favour cleaning the sensors every now and then over dealing with servo motors in these things, but it's a pretty interesting idea for sure
INTP:
I wonder what kind of garbage is so effusive that it slathers up a sensor from underneath.
Yes, what @wvmarle said. Waste bins are different throughout the world, but the dumpster ones in North America get lifted and tilted above the truck, so that everything falls out of the top of the dumpster. Any sensor therefore takes a pretty big hit every time, in terms of liquids, grime, organic products etc. Some of the worst sensors (not ones I made) I've seen had been on dumpsters for 1-2 years and were coated in a cake of 1-inch thick organic waste.
I'm used to garbage workers opening the container before it's tipped, to keep the lid out of the way of the falling rubbish... also I'm used to seeing that kind of containers being cleaned on a somewhat regular basis (just by hosing them down). So whatever sensor you use it'd have to be really robust!
That said, the ones at the waste collection point across the street from me are usually kept open anyway. Just as easy. Waste is there for no longer than a day (except Chinese New Year when garbage collection stops for a day) so it doesn't have the time to start rotting too much.
No moving parts? I think you're going to have to find a sensor that penetrates the thin layer of garbage but detect the actual garbage. IR might get through.
Although ment to work in water how about the transducer from a fishfinder?
Not up to date on exactly how they work but it is sonar which is sound waves sent through water.
Wont be as efficient but you are using it in a small enclosed area so it might work through the air instead.
wvmarle:
I'm used to garbage workers opening the container before it's tipped, to keep the lid out of the way of the falling rubbish... also I'm used to seeing that kind of containers being cleaned on a somewhat regular basis (just by hosing them down). So whatever sensor you use it'd have to be really robust!
That said, the ones at the waste collection point across the street from me are usually kept open anyway. Just as easy. Waste is there for no longer than a day (except Chinese New Year when garbage collection stops for a day) so it doesn't have the time to start rotting too much.
That happens here sometimes as well, though the norm is to just leave them closed when it's picked up. The containers are cleaned regularly with a strong water hose, so that's why I need this to be an IP67 product. It depends on the location how long the waste is inside of them; it can range for one day to up to 2 weeks.
Grumpy_Mike:
Use a lid like on a garden incinerator and house the sensor in the top to keep it out of the way of the rubbish.
Unfortunately, I want to design something that goes on your typical dumpster bin, like this:
so I have no control over the type of lid on them.
INTP:
No moving parts? I think you're going to have to find a sensor that penetrates the thin layer of garbage but detect the actual garbage. IR might get through.
You mean like the Sharp IR products? The issue I see however is that the composition of the waste material is highly variable, so it would be difficult to account for this and ignore it in the measurements.
Daz1712:
Although ment to work in water how about the transducer from a fishfinder?
Not up to date on exactly how they work but it is sonar which is sound waves sent through water.
Wont be as efficient but you are using it in a small enclosed area so it might work through the air instead.
Would these actually be successful in ignoring materials other than water though? I don't really have an issue with water for my ultrasonic sensors as it will just fall off or evaporate.
TolpuddleSartre:
Why not use a car parking sonar sensor?
They are designed for harsh environments.
I've looking at that as well, but the car parking transducers I find usually are pretty much identical the ones found in for example the JSN-SR04T, or the MaxBotix Sensors. So I'm expecting to see a pretty similar behaviour from them with regards to the contamination of the transducer.
Also on a more general note, here's an example of what my ultrasonic sensor may typically look like after a couple of weeks:
The red smudge on it was sufficient to block out the measurements completely.
So I'm expecting to see a pretty similar vehaviour from them with regards to the contamination of the transducer.
The ones on the front of my car are subjected to regular drenching (I live in the UK), extreme wind-chill and direct high-velocity insect, grit and stone strikes, and yet have worked flawlessly for the last five years.
I was just thinking of a basic distance sensor with a separate emitter and sensor. Spacing them might give you a beneficial blind spot of things that are too close due to n angle that's too steep for reflection.
TolpuddleSartre:
The ones on the front of my car are subjected to regular drenching (I live in the UK), extreme wind-chill and direct high-velocity insect, grit and stone strikes, and yet have worked flawlessly for the last five years.
Would you mean providing me a link (or an image)? I'm interested to delve deeper into this. It makes sense that a bit of dirt/grime wouldn't cripple a car right away.
INTP:
I was just thinking of a basic distance sensor with a separate emitter and sensor. Spacing them might give you a beneficial blind spot of things that are too close due to n angle that's too steep for reflection.
That makes sense. Though if some dirt has accumulated on the emitter (or receiver) like on the picture, do you think the signal could still actually carry through? How little of the signal needs to carry through for the receiver pick up on it?
Vitesze:
Would you mean providing me a link (or an image)? I'm interested to delve deeper into this. It makes sense that a bit of dirt/grime wouldn't cripple a car right away.
TolpuddleSartre:
You've never seen a car with parking sonar?
The ones I find look like they have the exact same housing for the transducer as the JSN-SR04T and MaxBotix sensors which I already have been testing. I was just wondering what specific parking sensor you're referring to here.
Grumpy_Mike:
I am from the UK and I have never seen a bin like that before.
Ahh...here in North America, this is really the standard dumpster that you will see almost everywhere at commercial sites. The dumpster is lifted using the two 'slots' on the sides, and emptied above the truck (we call it a Front-end Waste bin, as it's lifted over the front of the truck)
Henry Ford did not say, I cannot make a car, because there are no roads, and the rail-road is already in place.....
if you want to create a sensor that uses power and is on a truly harsh environment, you have to either change the rules of physics and the conservation of energy on the planet or just deal with the harsh realities.
in England, there is a small company called Premier,
they do waste removal, just like in the U.S. Premier Waste Hauling
and have the exact same sort of bins as in the U.S.
when the truck comes, the front gets all the rubbish across the lip and the HINGE is far from the debris.
You can simply add a bolt-on [weld on] device under the hinge.
the driver has a charger or pack of batteries.
the truck has the receiver
when the truck comes to lift the bin, it can read the sensor, the battery level, the whole bit.
the bit about Henry Ford is that he did not put up all manner of restrictions on how the car can be used, he certainly did not say that the tank had to last for 5 years without being re-filled. nor that it had to work without human intervention if it were covered in mud. there is a difference between design goals and arbitrary conditions. 99.9% of bins never get debris on the hinge part of the bin.
you can weld a 1 inch square tube on the hinge side of a container, create a 4 or 6 foot long set of sensors that slides into the tube for protection and then peeks through the open ports to read the bin level of the closest object.
a cardboard box or three from old refrigerators will make the bin appear to be completely full.
I missed that you probably want to use a cellular link, but do not want to pay $25 per month per bin for the cell link. with millions of bins, who would !
a LoRa or Wifi that searches for your trucks, and that means any and all haulers so equipped. would then report from the mesh, back to your central database and then inform the proper subscriber of the bin status.
that means if a W.M. (Waste Management ) truck so equipped, drives through the industrial park, it would pick up and read the status of all the Republic bins, the Rumpke bins and the Premier bins......
and when a ADC truck drives through the next day, it too would read all the bins and the truck would report back.
the Truck would be the cell link, not the bin.
only the truck company would get to see their bin's data. the rest would be sent back to your offices and linked to the respective clients.
battery changes would be done when the bin reports a low batt condition and the drivers would need to get out of the cab every so often.
The added collection pickups, the customer convenience, all of that would pay the extra 2 minutes the driver would spend, once a year or every 6 months for changing batteries. and if the bin is emptied and the sensors do not reprot the correct empty condition, the Driver would have to get out and wash off the sensors. which, you could auotmate at some point.
Remember, some brave soul said,. ' give me 10's of thousands of dollars for a special truck with forks and even more thousands for bins with welded pipes on the side so you can fire 4 guys and have one driver pick up trash....
now the standard, but that guy had some big 'cobblers' back in the day.
the simple, first test can be on a Uni-Strut that is welded to a bin, and you install a faceplate, that bolts on. later, you can have holes cut in square tubes and have them welded to the bins.
patent the idea, like yesterday, and get world wide patents before that !
there are companies whose goal is to copy patent applications and just use that to apply for patents in every country that was not already filed. the patent goes the first to file, not the creator. long gone is the day when you prove you had the original idea so were awarded the patent.
as for those who have car bumper parking sensors, you also have a 80 MPH 'car wash' every time it rains to keep the debris from piling up.
and the sensor is designed to ignore anything less then 20 cm.
