GRAPHICAL ABSTRACT
A Machine Learning (ML) strategy was used to enhance mid-infrared (MIR) laser reflectance spectra based detection of high explosives (HE) in mixes with soil samples. The Artificial Intelligence (AI) methodology termed “Self-Simulated Learning Artificial Intelligence” (SSLAI) demonstrated that it is possible to detect a chemical threat in solid matrices such as natural soil. There is an excellent possibility to transfer the methodology to field applications where the detection time is essential, and where often these solid matrices are unknown.