Help with designing a roof snow load sensor and app…

Thread Starter

jojjo

Joined Mar 7, 2024
4
Hello
I will develop an electronic product that can read the amount of snow/m2 on a house roof. Have intended to use a laser that is fixed in a stable place in the attic. on the truss thing there is a mirror on which the laser beams shine. when the tact chair bends, the laser senses it and should then be able to convert the deflection into kg/sq.m. the data obtained should be sent to an app that the owner can read. everything must happen in real time.
How to build such a product in the best way? I'm not that well versed in electronics but maybe someone can help me along the way.
Best regards
// Jojjo
 

Ya’akov

Joined Jan 27, 2019
9,236
Hello
I will develop an electronic product that can read the amount of snow/m2 on a house roof. Have intended to use a laser that is fixed in a stable place in the attic. on the truss thing there is a mirror on which the laser beams shine. when the tact chair bends, the laser senses it and should then be able to convert the deflection into kg/sq.m. the data obtained should be sent to an app that the owner can read. everything must happen in real time.
How to build such a product in the best way? I'm not that well versed in electronics but maybe someone can help me along the way.
Best regards
// Jojjo
Welcome to AAC.

You’ve made a basic error in defining your problem by bundling the solution into it. This makes the solution your problem even though, as you say, you are “not well versed in electronics”. From your description of the problem/solution are you are also not well versed in engineering in general.

Step back and begin again, isolating the problem from the solution, something like:

I want to design and build a device that can determine the depth of snow on a roof in kg per m² and display the result in realtime using an app.

It must operate on roofs with ??? style construction.
It must have a sensitivity of ???.
It must have an accuracy of ???.
The budget is ???.
The deadline is ???.

Frankly your laser deflection idea strike me as an unlikely solution for a variety of reasons. There are many ways to measure strain, which is what you are suggesting to do, that are more robust and already packaged.

If you can fill in the gaps, I am sure there are many people here who can and would be wiling to provide some help.

One more thing, the title of your post is pretty much useless. In the future please make your titles more informative about the main thrust of the post. In this case I have edited it for you.

Again, welcome to AAC and good luck with your project.
 

Thread Starter

jojjo

Joined Mar 7, 2024
4
I will try again and rewrite my question.
The problem today is we often guess how much snow/water weighs on our roofs. Which usually results in broken roof tiles, safety equipment etc.
Instead of measuring snow depth and measuring manually, I want to develop an electronic equipment that checks the weight/m2 continuously.
It should be easy for the home owner to obtain the snow weight in real time via e.g. App.
My idea is to develop an outfitting that can be mounted on different types of roof constructions (indoors). Since roof constructions are different and can handle different weights, the equipment should be programmable based on the conditions of individual roofs.
Cantilever trusses can vary between 10 -35 meters
Roof types can be flat, to inclined 35 degrees
When the rafters/beams approach their maximum limit, the home owner must see this well in advance and can plan for snow shoveling.
Measured values should be within 250/kg/sqm - 900 kg/sqm

I have a small budget to work with but thought I would ask if you have any good ideas about such equipment.
Best regards
 

Ya’akov

Joined Jan 27, 2019
9,236
It seems possible to instrument a roof so you can tell the load it is bearing, but:

  1. not with a single sensor
  2. not with high precision
  3. not for a reasonable cost

Your idea of using a laser to measure deflection is not crazy or stupid, but it is almost certainly impractical. The practical problems implementing it in such a way that it wasn’t a research and development project for each installation would almost certainly make it unaffordable.

It would also suffer from the same problems as other potential methods: you are proposing to use the roof structure for something it was not designed or built for, as the cantilever in a strain sensor.

1709888948887.jpeg
a typical strain sensor uses a strain gauge bonded to a cantilever
that can be the item being measured or a proxy and as the
cantilever flexes the strain gauge is stretched, changing its
electrical properties in a way that can be correlated to the strain
it is being placed under, which can be converted to weight

To have a calibrated sensor the behavior of the cantilever under load must be characterized so it can be applied to the calculation of strain. It doesn’t have to be linear, but it does have to be:

  1. sufficiently variable to be detected in the range of interest​
  2. repeatable​
  3. consistent in changing environmental conditions such as temperature and humidity​
  4. relatively unaffected by age and cycles of loading and unloading​

I am not convinced that wooden beams can fulfill these requirements.

You also have the problem of trying to use a single point measurement as a proxy for the entire roof. While research might show this to be possible, on its face is seems very unlikely to prove accurate. This means you will probably need an array of sensors.

My suggestion for an approach, this is, by no means, a solution—is to consider the idea that you could do this with a roofing system. This wouldn’t work for retrofitting, but for new construction you could develop a system that embeds many, low cost sensors into the roof as it is laid down.

These sensors could potentially even be wireless, though that brings its own troubles so it may not be a good thing. In any case by using a purpose built grid of cantilevers and gauges, you could do what you want.

One more thing: if you change your requirements from ”display kg/m² in real time” to “display green, yellow, and red indicators showing the danger level” you might be able to do it as a retrofit, though it will still require a lot of research concerning the behavior of roofs and support structures.
 

camerart

Joined Feb 25, 2013
3,731
Hello
I will develop an electronic product that can read the amount of snow/m2 on a house roof. Have intended to use a laser that is fixed in a stable place in the attic. on the truss thing there is a mirror on which the laser beams shine. when the tact chair bends, the laser senses it and should then be able to convert the deflection into kg/sq.m. the data obtained should be sent to an app that the owner can read. everything must happen in real time.
How to build such a product in the best way? I'm not that well versed in electronics but maybe someone can help me along the way.
Best regards
// Jojjo
Hi J,
For something like this, I usually go for the simplest cheapest method to test, and improve on it as I go.
My first thought is the roof can be it's own strain guage.
If you can fix something stable under it with a microswitch, and add an opposite button on the roof, as the roof sags, it will switch the microswitch, then you figure out what to do once it's contacted.
C.
 

Externet

Joined Nov 29, 2005
2,235
Install one single kitchen scale of known area above the roof and multiply its reading to reach/account for all the roof area.

Better if has a remote display and power switch.
1709926102256.png
 

Externet

Joined Nov 29, 2005
2,235
And an application ??? Come on ! Use a calculator. :rolleyes: Or want a video surveillance camera aimed to the display sending the image 24/7 via internet to the owner ? Or wireless serial data ? Someone is watching too much television !
 
Last edited:

Reloadron

Joined Jan 15, 2015
7,546
Personally I would install several strain gauges which will measure strain on the roof. The problem is weight will not tie back to snow depth. NYC born and raised and now living in Cleveland I know all about snow and ice
Early snows we see as heavy snow, it's wet but as winter deepens and we get the cold sub zero snow it's a much lighter powder. 24" or powder may not weigh as much as 3" of the heavy wet stuff. So now we want weight on roof and we also wasn't snow depth.

Ron.
 

Tonyr1084

Joined Sep 24, 2015
7,956
Step back and begin again, isolating the problem from the solution, something like:

I want to design and build a device that can determine the depth of snow on a roof in kg per m² and display the result in realtime using an app.
Snow can weigh differing in numbers. Area is another consideration. Angle of the roof is yet another. Then there's the material the roof is constructed of. Since wood can be stiffer or weaker depending on species and whether it's old growth or new growth, there's going to be a whole lot of variation that needs to be considered.
I want to develop an electronic equipment that checks the weight/m2 continuously.
Weight is the better approach. But those other factors are still a menace to a solution. But there IS a solution.
If it was possible it would have been done by now..
But there IS a solution. Just because it hasn't been done "Yet" doesn't mean it can't be done. Other factors involved is how much time and research does someone want to put into developing such a device. Then once ascertained, is it practical or impractical?
Install one single kitchen scale of known area above the roof and multiply its reading to reach/account for all the roof area.

Better if has a remote display and power switch.
View attachment 317205
I was thinking something along the same line, some kind of scale. Kitchen scales didn't come to mind but that offers one possibility. But again, the type of roof (materials such as steel, asphalt or slate, etc.) will redirect snow loads differently. I have a steel roof at a 2:12 pitch. Snow would tend to accumulate more on a low pitch roof, but on my steel roof the snow would creep over the edge. The exposed snow would turn to ice as it extended over the edge then give way and break loose and crush the garden (or anyone who happened to be beneath it when it collapsed). I have an interesting picture of this on an older computer so I can't just upload it, but I used to get these four foot arc's of snow hanging over the edge.

As mentioned, a 2:12 pitch will hold more snow than a 4:12, which is minimum standard requirements where I live in these days. I put solar panels on the roof and before the contractor could do so it had to be evaluated by an engineer who knows a whole lot more about roofs than I do. Given that extra weight PLUS the heavier wet snow has me concerned about a similar issue as the one you're trying to solve.

Going back to the scale suggestion, the area of the surface of the scale needs to be known as well as the area of the roof. But such things like Dormers can capture more snow than a flat open roof. In those areas wind blown snow will accumulate and contain more weight than areas where the snow has been blown away from. So where do you put your scale? Which is why someone before me said you'd need to sense the weight in several areas. Logistically it becomes an encumbered solution to a problem where if you shovel a heavy wet snow off the driveway then you know there's a lot of weight on the roof. A roof rake is a cheap alternative and is safer than going on a potentially slippery roof. And at steeper angles, being up there with all that snow, causing an avalanche while you're up there can easily drag you off the roof.

Perhaps a heated roof would be cheaper; just like the gutter tape they sell to melt the snow at the edge of the roof and in the gutters to keep ice dams from forming. But who wants to heat a roof every time it snows? Unless you can visually determine when it should be turned on and off.

Hate to be the guy who finds problems with the solution, but sometimes finding a solution means finding what doesn't work. Just ask Edison. They say he found 10,000 ways NOT to make a light bulb before he found the ONE way to make it. (back then)
 

Externet

Joined Nov 29, 2005
2,235
Hi Tony.
The 'owner' may have strong reasons for being after such data; but still does not make sense to find out if the roof has now 37.2 Kg of snow and it was 31.9 Kg six and a half minutes ago, and 11.3Kg of snow last midnight.
Being any sloped roof, a 'kitchen' scale can still tell the weight on it multiplying the reading. Being 'shaded' by a building next to it or a tree or with funneled wind does not give good results as there is poor uniformity. We know that and the 'owner' must be aware.

If past tense is allowed, weighing the water melt tells how much has decreased; 1 Kilogram = 1 litre. Still sounds more like a test to someone problem solving abilities than a real need.
 
Last edited:

Tonyr1084

Joined Sep 24, 2015
7,956
On the laptop, had to scroll through a ton of pictures, but I found them:
Glacier Roof 01.JPG

Glacier Roof 01.JPG

Glacier Roof 03.JPG

Two days later, just before it collapsed on its own:

Glacier Roof 02b.JPG
Until I put up snow dams this was the case. The letter carrier was at risk of injury should breaking the ice sickles might have caused a collapse. I used to get out there with a pole and punch holes in the berg until it collapsed. And it crashed with a thunderous boom.
 
Last edited:

Thread Starter

jojjo

Joined Mar 7, 2024
4
Hello
There was not much positive to read, only problems and obstacles. just want to inform you that the technology exists and where you read snow weights and send it in real time. As for sensors, you should determine where the most snow is normally deposited on the roof, place the sensor there and take a reading. Why bother with everything, it's always easy to find problems.
// Whoops
 

Reloadron

Joined Jan 15, 2015
7,546
Hello
There was not much positive to read, only problems and obstacles. just want to inform you that the technology exists and where you read snow weights and send it in real time. As for sensors, you should determine where the most snow is normally deposited on the roof, place the sensor there and take a reading. Why bother with everything, it's always easy to find problems.
// Whoops
More a matter of pointing out potential obstacles and problems to someone. Had you mentioned that you were well aware of everything covered people would not have mentioned them. That said I would use strain gauges to measure strain on the roof. To be more specific I would use 350 Ohm strain gauges with each a leg of a bridge. I would use Micro Measurements strain gauges because I have used them in the past. OK, that's all I got.

Ron
 

ThePanMan

Joined Mar 13, 2020
800
I think people pointed out the problems you would encounter because helping you design a "roof snow load sensor and app" has so many variables. I, too, like the scale approach. But second choice would be to measure the stress in the beams. But I didn't read where anyone mentioned wind loading. I don't want to find yet another problem for your solution but it's hard not to consider all the ramifications. Even using a strain gauge on the roof trusses has issues. Someone already mentioned different types of woods and such but there's also humidity and moisture that will cause the wood to respond differently.

So how would I (me personally) design a snow load sensor? Perhaps a wire stretched across the span of the trusses. As weight increases tension on the wire would increase and thereby change the reading from a strain sensor. One sensor for each roof truss to give input from multiple points, gather all the inputs into a single average weight, or maybe two or three zones of indication. Then via bluetooth or IoT connectivity to indicate on a receiver what is being measured. The hard part would be understanding the data you were receiving. How much is no problem, how much is a concern and how much is reaching a critical point prior to failure. Then whether you want to display a number or set of numbers, or if you want indicator lights, 3 green lights is a clear roof, 2 green and 1 yellow indicates there's a snow weight on the roof, 1 green and 2 yellows indicate building up snow, 3 yellows is building up a heavy wet snow, 2 yellows and one red indicating cause for concern and time to take action, 2 red and 1 yellow indicating a critical point where immediate action is required and all 3 reds indicate an emergency situation.

Now, at what point do I trigger each level?
 

Tonyr1084

Joined Sep 24, 2015
7,956
Someone already mentioned different types of woods and such but there's also humidity and moisture that will cause the wood to respond differently.
Humidity would definitely need to be considered. When humidity goes up wood expands. Your strain gauge would see that and interpret it as meaning something. Maybe it may mean you have a heavy snow load on the roof in the middle of July or August. Unless you live in the southern hemisphere, getting an alarm that the roof is about to collapse due to heavy snow in the summer time would inspire distrust of the system you've designed.

Good points Pete.
 

Ya’akov

Joined Jan 27, 2019
9,236
It should come as no surprise that commercially available roof snow load monitoring systems use beam deflection as a proxy for direct measurement of snow weight. It should also be no surprise that (it appears) there are no such systems for wooden structures.

As I specifically mentioned above it would be straightforward to use the structural beam as the mounting point for a strain gauge—but not the beams of a residential wooden truss as they wouldn’t have enough consistency to either be calibrated for kg/m² or for maximum safe deflection.

Using a bathroom or other scale doesn’t change that. You still need to know the meaning of what it reads out. The TS actually proposed monitoring beam deflection in the original post, but using a laser is a vague way. The idea could be made to work but it is a lot of extra effort to do something that has fundamental shortcomings.

In my followup (#6) I said:

To have a calibrated sensor the behavior of the cantilever under load must be characterized so it can be applied to the calculation of strain. It doesn’t have to be linear, but it does have to be:

  1. sufficiently variable to be detected in the range of interest​
  2. repeatable​
  3. consistent in changing environmental conditions such as temperature and humidity​
  4. relatively unaffected by age and cycles of loading and unloading​

I am not convinced that wooden beams can fulfill these requirements.
I still can’t find any evidence anyone has done this commercially, and I can see why. I also suggested that retrofitting a system has many more problems than doing one designed to be part of a particular roofing system. This would seem the most fruitful direction and offers many possibilities.

While it would certainly be possible to incorporate such a system into roof supporting structures designed for the purpose, it might even be possible to do something that could be installed under the roofing shingles, slates, or tiles for a roof being refurbished.
 

Externet

Joined Nov 29, 2005
2,235
Perhaps a wire stretched across the span of the trusses. As weight increases tension on the wire would increase and thereby change the reading from a strain sensor.
Hi. A wire will expand and contract with the extreme temperature changes in an attic making readings in error. The kitchen scale suggestion and multiplication does not even need to be on top of the roof. Can be in a place with same exposure to snow. Perhaps is also immune to humidity and temperature.

There was not much positive to read, only problems and obstacles. just want to inform you that the technology exists and where you read snow weights and send it in real time.
----> I would like to see the original poster showing the approach taken and results. So let's wait for a future disclosure.
 
Top