Some years ago, I went to a demo and presentation by a local UAV survey company that do a lot of agricultural work. They were using hyperspectral imaging to detect crop condition including specific diseases. That approach used resource-hungry post-processing but it was not difficult to see the potential, as technology marches forwards, for SAR.
Likewise the potential for military use. Being able to identify the location on a battlefield of clothing with materials made in Moscow and clothing with materials made in Minneapolis would have some advantages and there are varied opinions about whether that realtime detection technology exists at this time. Even if it does, a target identity system and a wide-scale SAR system are quite different beasts. It is important to note that this is very much a daylight tool that normally requires the full daylight spectrum to produce the expected signature. (It is not entirely inconceivable that a narrow spectrum programmable floodlight might one day make night searching feasible but that is yet another big tech step away.)
During the intervening years I have sought out manufacturers and academics who are knowledgeable in this area of research and manufacture. The academics tend to be astrophysicists who of course have been using this to tell us what distant stars and planets are made of for some decades. Like the agro-surveys, they have been using resource hungry post-processing and in some cases, rather like the military, they have been focused on a small target area. They make clear that the SAR task would be hampered by the sampling algorithm that would have to be used to bring the data load down to levels that could be dealt with in real-time. So that ends up with any early attempt at this tech in SAR probably having the same disadvantages as Mark One Eyeball: It will scan along certain paths but there will be gaps.
So what might we expect to detect that would make this a useful SAR tool? Well, at sea, the most useful might be the nylon facing materials used on lifejackets and liferafts. One might foresee a time when the IMO might dictate standards that make items easier to detect. Maybe even the Keratin in human hair, assuming that the misper is not bald. On land, most outdoor clothing is specific regions of the world will be made from a small number of compounds not found in the same natural environment.
So the task is to scan a wide area ahead of a search aircraft, pick out areas with an unusual spectral signature, check those against an ordered list of materials that might indicate human presence, calculate location, and alarm appropriately. Not nearly as simple as it sounds. All the bits exist to do this but putting it all together in a manner that produces a good SAR tool requires deep pockets and may well be some years away.
An American project claiming to be able to do all this was around a few years ago. However, it appears that although demos showed some potential, real searches proved too challenging. The fact that the companies involved seem to have disappeared tells a tale. The scanning and sampling rates available with current technology could easily have been the problem.
A few organisations presenting at the ICAR Air Commission have mentioned this technology so it's clear several of us are thinking along the same lines. It has been mentioned to UK MCA Aviation's technical assurance team in the preparatory meetings for their UKSAR2G and they are receptive to such new technology solutions.
Think big!
Ask stupid questions!