Leveraging AI to Enhance Olive Oil Quality, Efficiency and Trust

Over the past two years, arti­fi­cial intel­li­gence (AI) in the form of large lan­guage mod­els – such as Gemini, ChatGPT and Bard – has cap­tured the pub­lic zeit­geist.

However, arti­fi­cial intel­li­gence has been help­ing olive farm­ers improve effi­ciency and deci­sion-mak­ing for the past decade.

Earlier this year, city offi­cials in Jaén, Spain, gath­ered to dis­cuss the many roles arti­fi­cial intel­li­gence can fill for olive farm­ers and millers in the world’s most prodi­gious olive oil-pro­duc­ing province.

Among the atten­dees was Decidata Analytics, a firm spe­cial­iz­ing in arti­fi­cial intel­li­gence used to improve data ana­lyt­ics.

Project man­ager Javier Canego Martinez told Olive Oil Times that arti­fi­cial intel­li­gence could improve trace­abil­ity and help build trust in a sec­tor plagued by neg­a­tive head­lines about fraud in recent months.

“AI tools can help ensure that the olive oil is authen­tic and high qual­ity by check­ing the type of olives used,” he said. ​“This can be done by ana­lyz­ing pho­tos of olive pits using image recog­ni­tion tech­nol­ogy. By pro­vid­ing accu­rate infor­ma­tion about the prod­uct, com­pa­nies can build trust with con­sumers and main­tain a strong brand rep­u­ta­tion.”

Indeed, researchers from Italian uni­ver­si­ties have trained arti­fi­cial intel­li­gence to iden­tify the prove­nance of extra vir­gin olive oil using its phe­no­lic com­pounds and sterols. In a 2022 study, they reported that the AI cor­rectly iden­ti­fied locally pro­duced Taggiasca Ligure olive oils.

Along with ver­i­fy­ing authen­tic­ity, Canego Martinez believes AI-pow­ered tools can help pro­duc­ers and retail­ers tar­get cus­tomers with per­son­al­ized adver­tis­ing.

“By under­stand­ing what con­sumers like and how they behave, com­pa­nies can cus­tomize their mes­sag­ing and pro­mo­tions to reach spe­cific groups more effec­tively,” he said.

“AI can fur­ther pre­dict mar­ket trends for the olive oil sec­tor, along with con­sumer demands, help­ing com­pa­nies adjust their strate­gies ahead of time,” Canego Martinez added. ​“This allows them to opti­mize inven­tory, pric­ing and mar­ket­ing efforts to stay com­pet­i­tive.”

Diego Hueltes, a com­puter engi­neer and con­sul­tant, has demon­strated that AI mod­els can accu­rately pre­dict prices. He devel­oped three pre­dic­tive price mod­els using his­tor­i­cal pric­ing data, cli­mate data, and data about pre­vi­ous har­vests to pre­dict prices for 2017 and 2018 accu­rately.

Most of the AI tech­nol­ogy for these appli­ca­tions is not as well known as the con­sumer-fac­ing large lan­guage mod­els that yield the cur­rent gen­er­a­tion of gen­er­a­tive AI.

However, Canego Martinez said gen­er­a­tive AI can help any­one involved in the olive oil world fill out doc­u­ments and forms and pro­duce ​“high-qual­ity con­tent, includ­ing social media, blogs, videos, and other mar­ket­ing tools.”

“Olive oil pro­duc­ers are slowly begin­ning to see the advan­tages of mov­ing into this area,” he said. ​“They can see how AI can not only enhance brand­ing for olive oil com­pa­nies but also improve over­all cus­tomer sat­is­fac­tion.”

During the event in Jaén, offi­cials dis­cussed the use of AI for cus­tomer sup­port chat­bots and how farm­ers and millers can use exist­ing AI tools to pre­dict har­vest tim­ing and mar­ket trends.

Along with arti­fi­cial intel­li­gence, there was dis­cus­sion about using blockchain tech­nol­ogy to improve trace­abil­ity and crack down on theft and fraud.

In addi­tion to the mar­ket­ing and cus­tomer-fac­ing parts of the busi­ness, Canego Martinez said farm­ers and pro­duc­ers can use AI to make their oper­a­tions more effi­cient and sus­tain­able.

“AI allows for opti­miz­ing pro­duc­tion processes, mak­ing them more effi­cient and sus­tain­able,” Canego Martinez said. ​“This includes the appro­pri­ate use of nat­ural resources such as water and fer­til­iz­ers, reduc­ing envi­ron­men­tal impact, and pro­mot­ing respon­si­ble agri­cul­tural prac­tices.”

At a June con­fer­ence in Mengíbar, another city in the province of Jaén, researchers pre­sented an AI plat­form that accu­rately pre­dicted olive har­vest tim­ings a sea­son in advance using mete­o­ro­log­i­cal and his­toric har­vest data.

“AI and machine learn­ing also play a fun­da­men­tal role in objec­ti­fy­ing deci­sion-mak­ing,” Canego Martinez said. ​“These tech­nolo­gies allow com­pa­nies to ana­lyze large vol­umes of data and extract pre­cise, evi­dence-based insights. This elim­i­nates sub­jec­tiv­ity and human bias, enabling more informed and fair deci­sions.”

Canego Martinez also pointed to pest con­trol as an exam­ple of how AI tools can help olive farm­ers in their daily work.

“Machine learn­ing mod­els can help pre­dict when pests might show up by look­ing at past data and envi­ron­men­tal fac­tors, which is extremely use­ful for farm­ers when mak­ing crop man­age­ment choices,” he said.

“They can also give a heads-up on future crop yields based on how things are grow­ing right now, mak­ing it eas­ier for farm­ers to plan ahead,” Canego Martinez added. ​“Optimizing irri­ga­tion is also a pos­si­bil­ity, like what Agrow Analytics offers in Spain… Moving for­ward, the appli­ca­tion of AI and machine learn­ing will con­tinue to pro­duce higher-qual­ity oil.”

• El Blog