Recent Advances in Food Process Engineering

Food waste represents the largest single stream of avoidable organic loss in the global food system, and its conversion into high-value ingredients has become a central ambition of circular bioeconomy policy. Progress in the underlying process engineering has nonetheless been uneven. The purpose of this critical narrative review is to evaluate, rather than catalogue, the engineering strategies through which heterogeneous food-waste streams are converted into ingredients of nutritional, functional or material value, and to establish where the evidence supports confident conclusions and where it does not. Literature was identified through structured searching of open scholarly indexes and registries, supplemented by citation tracking and by institutional sources, with appraisal focused on demonstrated scale, transparency of mass and energy accounting, and the coherence between experimental performance and system-level assessment. Four themes organise the synthesis: feedstock composition and variability as a design constraint rather than an operating nuisance; separation-led recovery of structurally intact molecules, including solvent design, energy-assisted intensification and pressurised fluids; bioconversion-led routes through carboxylate, polymer, protein and peptide platforms; and the integration of these unit operations into cascades whose economics and environmental performance can be defended. The evidence is strongest for laboratory and bench demonstration of individual recovery steps and for the direction of economic sensitivity in modelled biorefineries, which consistently identifies feedstock scale, high-value co-product price and downstream purification as dominant. It is weakest precisely where circularity claims are made: sustained operation on real, variable waste, solvent and enzyme recycling over many cycles, and life-cycle accounting whose boundaries and allocation rules permit comparison between studies. Reported gains from intensified extraction and from cascade sequencing are rarely accompanied by the energy and solvent inventories needed to verify them. Confidence in current conclusions is therefore asymmetric, and the field’s principal deficit is not the discovery of new recovery chemistries but the demonstration of integrated, instrumented and reproducible processing at pilot scale under realistic feedstock variation.