Hacker News new | ask | show | jobs
by leaving 1241 days ago
This is mostly a data analysis job.

If the capsule is on or very near the road, a single trip from start to finish with a NaI(Tl) or CsI scintillation counter will find it. Every CBRN team in the world has these, so they have obviously completed the first run at this without finding it. The capsule is not on or near the road.

Australia is famous for road trains on the Great Northen Highway. They travel fast and heavy on the paved road that runs through the outback. If the tires of one of those hit the capsule, they could fling it a fair distance into the desert. There's nothing between the road and the sandy desert to stop it.

If the capsule is a fair distance off the road, the inverse square law is going to reduce gamma detections closer to background. There won't be a clear peak in the data. If the vehicle with the detector moves slowly, there will be a wide peak in the time dimension that may just look like a natural variation in the background. If the vehicle moves fast, it may miss the source completely.

It will be a matter of looking at the data to try to find a pattern that indicates where the capsule may be. This will be made more difficult becuase they don't likely have a pre-capsule-loss run that would allow them to subtract normal variations of the background.

Adding to the problem is the stochastic nature of radiation. It is entirely possible to have a significant peak in the count rates that is not a strong point source; just a random variation in the background. Lower peaks from random confluence of background sources are likely; higher peaks are less likely, but not impossible. It may require a large number of runs to eliminate random events.

Real vehicles on real roads experience a lot of variations in speed, plus random starts and stops because the driver had to pee, or eat, or because there was a dead kangaroo on the road or something. A fair bit of work will need to be done to standardize multiple runs.

This can be partly solved by using multiple detectors on the same vehicle, with the idea being that peaks that are the result of natural variations in the background should only affect one detector at a time--so they can be subtracted out. But, again, CBRN teams have enough resources that they have already done this and they haven't found it.

So it is reasonable to conclude that the capsule is either quite far off the road, or has been picked up by a passing vehicle.

Another possibility is that the idea that it "fell off a truck" is wrong and that some human grabbed it by dissasembling the apparatus. This could be out of curiosity, or ignorance, or a misguided attempt to steal the source or use it for nefarious purposes.

If that is the case, the capsule is going to be nearly impossible to find unless it happens to pass a fixed gamma detector somewhere, such as a port.

The incident is a good argument for building scintillation crystals into all cellphones. This would be cheap and easy. Small scintillators are not very sensitive, and they are not great for spectroscopy, but they are extremely cool, very small, don't interfere with the workings of the phone, and use extremely little power.

A network of scintillators in every phone would be amazing for finding stuff like this.

4 comments

You comment is fantastic because it demonstrates we are beyond finding it with a rolling road block and sensors on trucks. This is a recovery operation where you're going to want surveillance for when someone shows up somewhere (medical care provider) with radiation sickness and to start contact tracing. If a human has it in their possession, exposure is ongoing. If it's sitting in a ditch somewhere, maybe it's years before it's found, maybe it's never found (and perhaps buried if and when a torrential rain passes through the area, typically in the upcoming winter months).
>If the capsule is on or very near the road, a single trip from start to finish with a NaI(Tl) or CsI scintillation counter will find it. Every CBRN team in the world has these, so they have obviously completed the first run at this without finding it. The capsule is not on or near the road.

It was reported in the national news within 12 hours of being found lost, so I don't think they had done that yet, especially given the distance.

I know this is just a list of theories.

> or because there was a dead kangaroo on the road or something

There's few situations where someone might stop for a dead kangaroo, or really any other wildlife. Either because they see it's still alive, or because they're checking to see if it might be a female with live young.

If it's clear that it's dead, nobody is going to stop. On a road, in WA, in summer -- yeah you're going to smell that for a long ways. Scavangers will deal with it pretty quickly, anyway.

Do people ever eat roadkill kangaroos? I know it's fairly common for people to eat roadkill deer in America, if they know the kill is fresh.
Ever? Possibly. I've not heard of it though.

Some Aboriginal communities will hunt and eat Kangaroo.

Hunting Kangaroo for sport/'population control' is a thing in some areas. But the meat is usually either wasted or fed to animals. I understand that with wild Kangaroo you have to worry about parisites.

You can get Kangaroo meat at the supermarket though.

A network of scintillators in every phone would be amazing for finding stuff like this.

That was my first thought too. Do you think it’s physically possible to make a MEMS scintiallation device?

If you're going to put increasingly obscure health sensors in your phone, I think air quality would come before radiation.
No. MEMS means electromechanical, and scintillation needs solid state: a crystal and a photon detector of some kind. The crystal can't be miniaturized because there's a minimum thickness you need to capture the photon and Compton electrons*.
You're right, I stand corrected.

This page [0] has an image of a small scintillator plus photomultiplier, but probably too thick for a phone.

[0]https://scan-electronics.com/en/dosimeters/radiacode-101

Although some weird Japanese phone makers did have smartphones with radiation detector built-in (SHARP 107sh and 205sh), I don't think it would be a popular feature.

There isn't so many cases in daily life that one can encounter a radioactive source. Even if you are actively looking for one (don't do it), you will have a hard time if you don't work with radioactive sources occupationally.

Also, it's not physically possible to create a small and yet sensitive radiation detector.

There are radiation detection apps for cell phones that use the camera. Not very sensitive, but no new hardware is required.