I don’t know where or what happened with this machine, as he was the technical lead for his company and they were a sub-contractor on the machine. But it was supposed to be a lot more sensitive accurate and smaller than what they had at the time.

From my understanding they completed their part of the contract and never heard anything else about it.

I think it is an unsolvable problem. There's nothing particular about explosives that makes them easy to discriminate from non-explosives.

I'd love to be proven wrong, though.

The vast majority of explosives can be quickly differentiated from non-explosives by gas chromatography-mass spectrometry.

The main practical impediment to detecting concealed explosive material is that some of the explosive substance needs to pass through the GC-MS to be detected. So an explosive with extremely low vapor pressure (e.g. pure PETN) will not be readily detected this way. Plastic explosives made from PETN are more readily detectable, because the plasticizers are more volatile than PETN itself. GC-MS also won't work if the explosive is hermetically sealed so that none of it can reach the detector.

There are also methods based on e.g. terahertz spectroscopy, but AFAIK GC-MS is the most specific and common method of identifying explosive materials.

They actually test for both components of binary explosives in the US TSA. Source: a TSA agent, while he was going through all of my things because I guess I touched a cleaning product with ammonia in it or something.