What sensors/methods can be used from the roadside to determine if a truck is loaded or empty - without weighing it?

What sensor, using what physical principle, can penetrate a closed rectangular object like the trailer of a truck and tell if the space is empty or loaded?

Why is this question relevant? One of the biggest problems of maintaining & building roads is to understand how much load is transported over that road per time increment. A loaded truck can bring much more damage to a road than many cars can (a 18 wheeler can have the impact of over nine thousand cars). Current in-road, at-speed weighing technology can cost millions. The reason for this questions is trying to find a way to use existing low cost sensors (or a combination of multiple sensors) at the roadside to assist with estimating the weight of a truck. If we know the truck/ trailer combination is empty or full, we typically know the lowest and maximum load weight. And then can estimate the average load on this road per time increment. This is much better than current vehicle counts and is commercially very interesting.

Many thanks for staying on topic and trying to point me in the right direction.

Initial thoughts: It is easy if you can visually see the load (like on a logging truck). It is hard if it is a closed curtainsider, bulk load or container truck from a stationary roadside measurement unit.

Computer vision could be used to compare the bounce of the cab, analyze the suspension; use thermal camera to detect higher temperature of tyres or axis; acoustics of empty vs loaded. What other sensors/ physical principles should be tried like measure on an uphill stretch: slower truck of same vehicle class will be loaded truck.

Again, many thanks for sharing your thoughts on what physical principle could be tried to tackle the problem to detect if a confined space is empty or filled, from the outside without weighing it. An approximation with an acceptable error rate is fine.

• Mostly impossible besides putting sensors in the pavement. Mar 5 '20 at 19:29
• At some border crossings & customs posts x-ray scanners are used visualize the contents of closed truck containers. The trucks are stationary while they are zapped with radiation. I don't know of any device that can quickly assess a moving truck trailer, particularly if it is a steel shipping container.
– Fred
Mar 5 '20 at 21:30
• Consider instrumenting bridges and overpasses with strain gauges and deflection monitors. And catalog actual trailer IDs. You want to be able to compare costs/budgets against actual tax revenue to keep things in proper proportion. Bridges are a natural site for this sort of instrumentation. Mar 5 '20 at 23:55

Ultra sound sensors could be used, you know the general size of all available container sizes and the material it is made of create a lookup table for such empty containers returning sounds. sound returned from a loaded truck will be different. It will not give you mg though.

• So a truck full of feathers will be empty when your table suggests it should weigh several tonnes more than it does? Or how do you intend to quantify all haul-able things? Mar 5 '20 at 19:31
• as per the requirement, this case is out of scope. heavenly loaded truck is the requirement. Mar 6 '20 at 11:52
• So a Truck heavily loaded with sand and a truck heavily loaded with osmium, you intend to tell the difference how? Bear in mind you can only measure density. Not to mention you’re assuming a truck full of feathers is anything but light. Mar 6 '20 at 11:54
• Indeed it is density and packing factor, as per general accepted transport requirement, I would intend not to pack the container with feathers that would destroy its shape and property, which will not be denser than Osmium. "questions is trying to find a way to use existing low cost sensors ... at the roadside to assist with estimating the weight of a truck ... empty or full" .Requirement is have a 0 or 1 answer and not distinguish between material inside, which could have been achieved by some EM technology. Mar 6 '20 at 19:52
• The point is, what is considered for road damage is not simply 1 or 0 rather many things in between and ultra sounding all possible combinations to determine the threshold is simply prohibitive. Mar 6 '20 at 20:04

Measure the deflection of the suspension springs in the trailer. The greater the load, the lower the trailer will sit. You prepare a calibration chart using known weights loaded on a given brand of trailer and repeat for the most popular brands. Then you can detect how heavily the trailer is loaded by how much closer to the ground it is than it would be if empty.

The other way to do this is to measure the area of the tire's contact pad on the pavement surface, and measure the tire pressure. From these you can solve explicitly for the load that that tire is carrying.

• Will air-springs not give constant ride height? I was thinking the same thing about the tyre compression but decided it would be problematic as there are 2-wheels/axle and 4-wheels/axle configurations. Mar 6 '20 at 17:34
• I do not know! -NN Mar 6 '20 at 18:50

For multiple "spots" on a road, train an overhead, high-speed (maybe), sensitive, thermal imaging camera on the road. All else being equal, higher pressure from a vehicle passing over a "spot" will cause it to heat up more from compression, deformation, and destruction. The material should be briefly hotter along the "heat tire tracks" after a vehicle passes. The speed of the vehicle, the width of the tracks, and an estimate of the surface area of the tires on the road should be able to give you the info you need. It would also be a good idea to generate a curve on the change characteristics based on starting temperature. Cold pavement might show more or less change than hot pavement.

I have no idea if this idea is feasible with current IR camera sensitivity. Likely to not work well over wet pavement, over snow, and maybe even in fog.

I got pulled over and was issued a speeding ticket fifty-some years ago in the middle of nowhere, and there were no guns then to clock you.

I asked the cop how he estimated my speed, which by the way was pretty accurate?

He said he positioned himself after a certain curve on the road and used a vertical road sign as a plumb, and by checking the degree of the cars lateral tilt due to the centripetal force he could tell.

You can put a camera on a similar position, the loaded trucks will sway more. One can establish a cross table of loads and tilts for different trucks.

Very practical and almost no maintenance.

• Will air-springs counteract this? Mar 6 '20 at 17:36
• How would your setup look like? How do you calibrate the multiple trailer combinations? Many thanks for suggesting more practical detail. Mar 14 '20 at 3:58
• I edit my answer. To add detail. Mar 14 '20 at 4:11

Roughly tell between full and empty by measuring inertia. If the road surface is slightly twisted, then the truck will sway left and right a bit. Even with different suspension systems I can tell the difference when a truck drives over a particular stretch of road. Not sure how to automate detection.

• Putting a small obstacle on the road and then determine the difference between empty & loaded truck could work. But this needs an intervention on the road. It would be excellent to find an approach which is 60%+ accurate from the side of the road. Mar 14 '20 at 3:47