(Organic) Food: Fewer Losses through Sensory Detection

In the future, it should be pos­si­ble to mon­i­tor and assure the qual­i­ty of (organ­ic) food along the val­ue chain more quick­ly and effi­cient­ly than before. To this end, new sen­so­ry detec­tion meth­ods are being devel­oped that will help to reduce food loss­es. From har­vest­ing to logis­tics and stor­age to pro­cess­ing or the way to retail, meth­ods are to be pro­vid­ed with which food can be planned, dis­trib­uted and fur­ther processed in the best pos­si­ble way, tak­ing into account indi­vid­ual qual­i­ty char­ac­ter­is­tics. They are being devel­oped in the joint project “SHIELD — Safe domes­tic (organ­ic) food through sen­so­ry detec­tion meth­ods”, which is being fund­ed by the Bavar­i­an Research Foun­da­tion with €1.2 mil­lion. Coor­di­nat­ed by the Fraun­hofer Insti­tute for Process Engi­neer­ing and Pack­ag­ing IVV, this dig­i­tal trans­for­ma­tion of ana­lyt­ics, sen­sor tech­nol­o­gy and detec­tion in food qual­i­ty man­age­ment will start on July 1, 2021.

Togeth­er with sci­en­tif­ic insti­tu­tions and appli­ca­tion part­ners, fun­da­men­tal con­cepts are being devel­oped that also involve mon­i­tor­ing and con­trol author­i­ties such as the Bavar­i­an State Office for Health and Food Safe­ty. The pro­por­tion of organ­i­cal­ly farmed land in Bavaria has been ris­ing steadi­ly since 1994 and is to be increased to 30% of agri­cul­tur­al land by 2030 through the BioRe­gio 2030 pro­gram. At the EU lev­el, more­over, the area to be farmed organ­i­cal­ly will be expand­ed to 25% from 2023 under the new Com­mon Agri­cul­tur­al Pol­i­cy (CAP) and the Farm-to-Fork strat­e­gy. “Strength­en­ing qual­i­ty and safe­ty across organ­ic val­ue chains is there­fore also of great impor­tance for the com­pe­ti­tion of the Bavar­i­an organ­ic sec­tor. In order to reduce the need to sort out food raw mate­ri­als that are dam­aged or unsuit­able for human con­sump­tion, we need sim­ple and reli­ably man­age­able process­es that can be used to ana­lyze raw mate­ri­als as soon as they are received,” explains Prof. Dr. Andrea Büt­tner, Exec­u­tive Direc­tor of the Fraun­hofer IVV and joint spokesper­son for the SHIELD project.

Handheld devices and smart software facilitate planning

Sen­so­ry detec­tion meth­ods should make it eas­i­er to gen­er­ate qual­i­ty fore­casts and opti­mize logis­tics chains in order to address the actu­al needs of the food indus­try and con­sumers. By accu­rate­ly deter­min­ing the (organ­ic) raw mate­r­i­al qual­i­ty for pro­cess­ing or deliv­ery to retail, food loss­es can be sig­nif­i­cant­ly reduced.

“Among oth­er things, we are com­bin­ing sen­sor tech­nolo­gies, opti­cal meth­ods and smart algo­rithms to devel­op hand­held devices and smart soft­ware that can be used in small-scale oper­a­tions. In addi­tion, detec­tion meth­ods for the authen­tic­i­ty of both raw mate­ri­als and pro­duced foods are being established.”

- Dr. Susann Vier­bauch of the Fraun­hofer IVV

At the Chair of Aro­ma and Odor Research at the Friedrich Alexan­der Uni­ver­si­ty of Erlan­gen-Nurem­berg (FAU), mark­er sub­stances are defined from authen­tic sam­ple mate­r­i­al for qual­i­ty deter­mi­na­tion. These form the basis for the devel­op­ment of rapid meth­ods for incom­ing goods. The Fraun­hofer IVV, the Fraun­hofer Research Insti­tu­tion for Microsys­tems and Sol­id State Tech­nolo­gies EMFT and the Chair of Ana­lyt­i­cal Food Chem­istry at the Tech­ni­cal Uni­ver­si­ty of Munich are respon­si­ble for the eval­u­a­tion. Pro­fil­ing meth­ods are being devel­oped at the Chair of Food Chem­istry at FAU in order to be able to sys­tem­at­i­cal­ly detect unex­pect­ed problems.

In addi­tion to record­ing the qual­i­ty and safe­ty of the deliv­ered raw mate­ri­als, the rapid and pro­fil­ing meth­ods are also used to deter­mine the authen­tic­i­ty of the raw mate­ri­als. Dur­ing the project, the rapid meth­ods will be con­tin­u­ous­ly com­pared with fur­ther find­ings and val­i­dat­ed in prac­tice at the indus­tri­al part­ners. These find­ings are incor­po­rat­ed into a data plat­form that is used for com­bi­na­to­r­i­al analy­sis and cor­re­la­tion of the mea­sure­ment results.

Digital incoming goods and online logistics

The eval­u­at­ed low-cost sen­sors, which are used on site by the indus­try part­ners dur­ing goods receipt as a quick method as well as dur­ing stor­age, pro­vide valu­able data on the qual­i­ty, safe­ty and authen­tic­i­ty of the raw goods. Thus, the dig­i­ti­za­tion of incom­ing goods via NIR and opti­cal meth­ods gen­er­ates an enor­mous amount of infor­ma­tion, e.g., about shape, col­or, size and tex­ture. This infor­ma­tion enables reli­able shelf life pre­dic­tions through pat­tern recog­ni­tion and thus improves logis­tics and for­mu­la­tion dur­ing pro­cess­ing. At the Chair of Machine Learn­ing and Data Ana­lyt­ics at FAU, meth­ods and mod­els for shelf life and sales fore­casts are being devel­oped for this purpose.

This takes into account the fact that prod­uct-spe­cif­ic knowl­edge can be inte­grat­ed into the fore­casts. In this way, sea­son­al fluc­tu­a­tions in sales can be tak­en into account. Based on this, math­e­mat­i­cal opti­miza­tion mod­els are pro­vid­ed at the Chair of Busi­ness Math­e­mat­ics to sup­port the for­mu­la­tion process. These auto­mat­i­cal­ly gen­er­ate sug­ges­tions for for­ward-look­ing, sus­tain­able pro­duc­tion plan­ning. An online approach ensures that logis­tics and pro­duc­tion are always in an opti­mal state through con­stant reop­ti­miza­tion when new infor­ma­tion arrives.

As part of the project, a needs analy­sis is being car­ried out and knowl­edge trans­fer estab­lished at the Tech­ni­cal Uni­ver­si­ty of Nurem­berg at the Neu­markt i.d.Opf. site in the research area of man­age­ment in the organ­ic sec­tor. The aim is, in addi­tion to record­ing the over­all sit­u­a­tion in the organ­ic sec­tor, to define the spe­cif­ic chal­lenges and imple­ment the find­ings from the SHIELD project. In addi­tion, these are then to be trans­ferred to oth­er relat­ed prod­uct levels.