Will agriculture find its salvation in mathematics ?

No. It’s to come back to it and to open up spaces for reflec­tion. The “food sys­tem” is at the heart of dif­ferent pro­cesses : dis­tri­bu­tion, consump­tion, and pro­duc­tion. Pro­duc­tion used to be a direct func­tion of three fac­tors : soil, ani­mals, and humans. All of these have been made more pro­duc­tive with mecha­ni­sa­tion and nitrates (i.e. hydro­car­bons). In a way, we are eating oil ! The ques­tion today is : can we main­tain this pro­duc­ti­vi­ty without oil ? And it is very dif­fi­cult to ans­wer if we look at pro­duc­tion alone.

If we think in terms of the farm itself, one pos­sible ans­wer is low tech ; a return to prac­tices that exis­ted before mecha­ni­sa­tion. But the ques­tion of pro­duc­ti­vi­ty soon comes up, and with it the ques­tion of cost.

Ano­ther res­ponse is to increase tech­no­lo­gi­cal inten­si­ty. But this still poses many pro­blems : it is not easy to extract bee­troot with a robot. And ener­gy remains the key issue, even if we can ima­gine a shift to electricity.

On the other hand, if we broa­den our thin­king from the farm to its wider context, and in par­ti­cu­lar the food sys­tem, other pos­si­bi­li­ties appear. We can opti­mise orga­ni­sa­tion, making sure that agri­cul­ture is bet­ter inte­gra­ted into food and finance chains. There is a lot of poten­tial here because we are tal­king about a poor­ly orga­ni­sed, frag­men­ted and poor­ly model­led world, where many deci­sions are taken in a non-coope­ra­tive way.

To opti­mise the way things are orga­ni­sed, public poli­cies are nee­ded. The issue of food secu­ri­ty has been an invi­sible pro­blem in the public are­na for seve­ral decades, but just because the pro­blem has been sol­ved does not mean that it will not arise again. Ques­tions of food sove­rei­gn­ty will come up again soon, as cli­mate change will put a strain on pro­duc­tion systems.

Howe­ver, it is cer­tain­ly not a ques­tion of cen­tra­li­sing eve­ry­thing as we saw with Gos­plan, the disas­trous results of which are well known in agri­cul­ture. Rather, the chal­lenge is to bring about bet­ter coor­di­na­tion bet­ween players whose inter­ests are not cur­rent­ly aligned.

Hence, one of the hori­zons is a plat­form for Euro­pean food, with a cer­tain amount of data sha­red bet­ween all the players within the next thir­ty years. The plat­form will sup­ply the major dis­tri­bu­tors and will also allow finan­ciers to car­ry out risk ana­ly­sis. Plat­for­mi­sa­tion allows the mathe­ma­ti­sa­tion of agri­cul­ture. It is the key to bet­ter orga­ni­sa­tion. The chal­lenge is to make the dif­ferent chains (pro­duc­tion, dis­tri­bu­tion, finan­cing) more col­la­bo­ra­tive and to have new deci­sion-making tools. To do this, it is impor­tant to model these chains from end to end, from the farm to dis­tri­bu­tion and invest­ment funds, and to equip the deci­sion-making pro­cesses at all geo­gra­phi­cal levels. We need to put maths at the ser­vice of agriculture.

Yes, in the sense that the coope­ra­tive model in Europe com­pen­sa­ted for the fact that farms were too small by making it pos­sible, for example, to pool equip­ment : moder­ni­sa­tion and coope­ra­tion went hand in hand and a new level was rea­ched. These coope­ra­tives, some of which have become very power­ful, were orga­ni­sed as buying and sel­ling groups, with some advice.

But the vir­tuous side of this model was under­mi­ned by two phe­no­me­na : the first is that it belongs only to the far­mers. The second is that Euro­pean agri­cul­ture has been orga­ni­sed in silos, and that it has been orga­ni­sed within the fra­me­work of an agri­cul­tu­ral pro­duc­tion poli­cy. Public poli­cies have thus crea­ted a clear sepa­ra­tion bet­ween pro­duc­tion and food, which can still be seen today in the Green Deal.

It is time to recon­cile them, and mathe­ma­ti­cal models are capable of doing so. Plat­for­mi­sa­tion and model­ling offer a way of mana­ging col­lec­tive deci­sions and intro­du­cing more ratio­na­li­ty into them.

The agri-food sec­tor opti­mises its logis­tics and indus­trial pro­cesses and fore­casts its food sales using models. But nothing is coordinated.

Yes, but they too are mar­ked by an incre­dible frag­men­ta­tion. Agri­cul­tu­ral exper­tise is sto­red in thou­sands of small tools, spread­sheets, mini-simu­la­tors, small cal­cu­la­tors, deve­lo­ped ad hoc by far­mers, tech­ni­cal ins­ti­tutes, asso­cia­tions, agri-food manu­fac­tu­rers, coope­ra­tives or labo­ra­to­ries, without any conso­li­da­tion. These tools are gene­ral­ly “out­side the infor­ma­tion sys­tem”, i.e. they are not fed into any recur­rent data flow. They are also not very user-friend­ly, due to a lack of invest­ment, and are most­ly unu­sed. Last but not least, what they lack is a sys­te­mic approach.

On the tech­ni­cal side, plant and ani­mal gene­tics are model­led, as well as fer­ti­li­ser and plant pro­tec­tion pro­duct pres­crip­tions. On the finan­cial and agri­cul­tu­ral accoun­ting side, risk scores, per­for­mance indi­ca­tors and busi­ness plans are model­led because of regu­la­to­ry models. But agri­cul­ture has no sys­te­mic model at the level of indi­vi­dual farms. Deci­sion sup­port tools gene­ral­ly model only one facet of the living world : the control of a par­ti­cu­lar disease, the imple­men­ta­tion of a par­ti­cu­lar prac­tice, in short, a very small part of the ove­rall system.

The agri-food sec­tor opti­mises its logis­tics and indus­trial pro­cesses and fore­casts its food sales using models. But nothing is coordinated.

As for the models used by public poli­cies, they are obso­lete and serve to dis­tri­bute sub­si­dies. They are models of eco­no­mic flows and balances that look to the past and ignore the agro­no­mic dimen­sion. We are thus depri­ved of the tools to steer the Green Deal, which explains why ideo­lo­gy takes pre­ce­dence over rea­li­ty : the figure of 30% less pes­ti­cides is thus a poli­ti­cal figure, which is not sup­por­ted by data. In terms of data and model­ling, eve­ry­thing remains to be done.

It leaves us no choice. The pro­blems of poor coor­di­na­tion, frag­men­ta­tion and the sepa­ra­tion of pro­duc­tion and food lar­ge­ly explain the condi­tion of Euro­pean agri­cul­ture. It is a sec­tor that is poor­ly finan­ced by the pri­vate sec­tor and sup­por­ted by public funds, which is won­de­ring about its future and fin­ding it hard to invest.

Yet the real pro­blems are ahead of us : food secu­ri­ty, agro-eco­lo­gy, decar­bo­ni­sa­tion, soil res­to­ra­tion. Eve­ry­thing needs to be done, in a context mar­ked by cli­mate change, ten­sions over raw mate­rials and pro­bable tur­bu­lence on the world mar­kets for agri­cul­tu­ral and food products.

Agri­cul­ture, which had been sim­pli­fied at the cost of car­bon and sub­si­dies, has sud­den­ly become what it was : a com­plex acti­vi­ty, because living things are com­plex. And food is a very com­pli­ca­ted sub­ject. The sec­tor now has to deal with contra­dic­to­ry injunc­tions that place it on the thre­shold of a major disruption.

From the new geo­po­li­ti­cal situa­tion to the rise in the price of ener­gy and the­re­fore of inputs, all the ele­ments are present for a cri­sis, with domi­no effects. Even a minor ele­ment such as envi­ron­men­tal label­ling contri­butes to des­ta­bi­li­sing the system.

This major dis­rup­tion opens up a field for crea­ting new models, and the tech­no­lo­gy is there : without wai­ting for plat­for­mi­sa­tion, we are now capable of crea­ting and run­ning models that are sophis­ti­ca­ted and rich enough in data to allow mathe­ma­ti­sa­tion of pro­duc­tion, but also of consu­mer beha­viour and finan­cing – this last issue being cru­cial to acce­le­rate the transformation.

If we need to model, it is because we are col­lec­ti­ve­ly blind, and we look to the past when the chal­lenges are ahead. The actors are aware of what awaits them. The ques­tion for them is now to orga­nise them­selves to avoid taking the wave head-on.

Interview by Richard Robert