E-Agriculture

Ajit Maru

This member participated in the following Forums

Ajit Maru
Ajit MaruIndependent ConsultantIndia

If the "plot manager" makes the decision aboy the plot, it would address the gender inequity we all would like to address. We did this in with milch animals in dairy milk cooperatives in India. 

My first question was also in context of algorithms and artificial intelligence that will surely be used to make farm decisions in the near future. Even now, in the GeoAgri Gujarat project I mentioned previously, we anticipate making suggestions and recommendations to farmers. Since the same suggestions and recommendation would also be available for review, for example by Insurance providers, whose farming decision would be legally considered for an insurance claim? Would the GeoAgri Project and the Universities backing it be liable to a "malpractice" suit? These are real questions that are now making it difficult to implement the project?

I fully agree that many of us (as seen even in this e-consultation) fail to see "the current disconnected and cash based reality that dooms them to informal and non-transparent supply chain relationships" and carried away more with emotions around perceived "ownership" rather than use and usefulness.

 

Ajit Maru
Ajit MaruIndependent ConsultantIndia

The blog is an interesting interpretation of GDPR for research, and more so in the context of ARD. One possible gain would be that it can force a discussion in FAO, CGIAR and other such Institutions on use of data which these Institutions have been sidestepping so far.

(By the way, I could not post a comment on the website as the Captcha was not there/did not work). Also interestingly it asked for my name, e-mail and website!

Ajit Maru
Ajit MaruIndependent ConsultantIndia

As one of the initiators of this e-consultation, I would also like to offer, IMHO, some of its possible conclusions.

The proceedings of this e-consultation point that policy, ethics and law will need to be associated and considered with in conjunction with issues such as expressed goals and objectives of a society, its choice of technology, development of structures, infrastructure and capacities to understand where the discourse on data driven, information rich and knowledge intensive farming, agriculture and Agri-food systems should go ahead with. Without these components included in any scenarios, the discussion, both for policy and other components of a data ecosystem such as technology remains abstract.

Data driven agriculture, in most parts of the world where smallholder farmers are in a majority, is very much in its infancy. The concept has not yet percolated to the policy makers who really count.  The farmers who would be most impacted are not yet adequately aware of the consequences of this development (or paradigm shift in my opinion). There may be a few stray statements by National leaders around this new agriculture (such as by the Indian Prime Minister) but as they say “One swallow does not make it a summer”. And, with great dismay, I would like to state that we who participate in this discourse are not yet powerful enough to impact for change. The potential of technology is far ahead than what we discuss now. However, this does not mean we should not continue exploring this area. In fact, it is our duty to do so continuosly and lead the discourse. I speak from personal experience. Many of the major Institutions I have been associated with rejected the very ideas in this area which they now proudly call theirs when they were first presented.

I am increasingly convinced that this discourse should primarily be centered around what will bring greater efficiency, better economics, higher quality, safety and transparency in the entire Agri-food systems and then consider what benefits individual actors such as farmers.

I have been concerned with issues related to smallholder farmers in which I also include farm laborers since now about 50 years. I do not have any romantic notions about their livelihoods and quality of life.  I know that they face an existential threat with this new farming and agriculture and most probably most will not survive in many parts of the world in the next 20 years. I am almost sure for India, which has among the largest numbers who we call smallholder farmers, many who not necessarily own land (we in India call them “landless” farmers), that this is happening.  Our discourse should consider this and see how we can buffer this shift in agriculture for these farmers. I am convinced that in the short term of 20-25 years, data driven agriculture can ease these farmers to leave their traditional vocation with greater dignity. We must consider this within the ambit of the discourse I have described above.

We had a very enlightening e-consultation. Some of us will comb it and thread the lessons we all are learning to continue this conversation elsewhere.

My sincerest thanks to all of you to have participated.

Ajit Maru
Ajit MaruIndependent ConsultantIndia

Let me try and answer your two questions.

1. How did we get here?

In my opinion we are in the very early phase of what Thomas Kuhn called a "Paradigm Shift" in agriculture and Agri-food Systems. We are shifting from Agriculture 3.0 to Agriculture 4.0 (See Internet for more information on Agriculture 4.0). Documenting and looking at the past will not help charting the future when there is a paradigm shift.

2. What do we already know?

In my opinion, as we go through this e-consultation and from my previous experience, very little in the context of this paradigm shift.  At the moment, most of what is stated are opinions, many a times romaticised. For example, of holding a view that a smallholder wants to continue to farm and would want his/her future progeny also to make farming their livelihoods. 

Ajit Maru
Ajit MaruIndependent ConsultantIndia

One key question that we must now consider, in view of the developments in the use of data and information such as AI, Big Data and Blockchain and in turn in farming is "Who is the farmer?" Is he/she who owns the land, makes the decision for the farm and farming or "owns" the data of the farm?

Another key question is (and going against the original grain of this e-consultation) should issues be only discussed around benefits of using data and information for the "farmer" whoever he/she/it may be?  Or how it brings greater efficiency, better economics, higher quality, safety and transparency in the entire Agri-food systems?

Ajit Maru
Ajit MaruIndependent ConsultantIndia

Most laws view property as being tangible. When we try to frame those laws for intangibles such as data and information, we tangle ourselves. In my opinion (and as opined earlier) in the consultation, laws should be framed around use of data and tangible benefits accrued to the user along with this declaration of "rights" and their infringement may be the most appropriate legal approach.

As regards farmers' data let us first make it clear that an individual farmer's data is not of much value beyond his/her use. The value emerges from aggregation. Thus, to me, this should be termed and considered as a community's intellectual property right.  Since it is "right", its practical implementation is in its infringement. It will be the Sovereign/States' power to judge and punish the infringement. However, going by current trends of data and information shared,exchanged and used internationally, when let us consider infringement is outside the realm of the State's power there will be a need and International arrangement, agreement or treaty with a functional regulatory mechanism that will be needed to judge and punish these types of infringements.

In my opinion, we now need an International Treaty on flow, sharing and exchange of data and information related to agriculture, food and nutrition. The ITPGR is inadequate in its provisions to deal with data other than that related to genetic resources. Also it has been found practically difficult to act against infringements and for that reason many countries are not cooperating in amending the treaty to include more genetic resources. Along with international treaties, we must have more robust regulatory mechanisms possible like the WTO. 

The other alternative, not entirely out of the box, would be to form Universal human collectives eliminating individual property ;-). Some very famous philosophers have already examined this idea.

Ajit Maru
Ajit MaruIndependent ConsultantIndia

Hugo, you are on track on discuss business models.

Bear with me for a somewhat long text to explain.

The area of economics of agricultural information has somehow evaded scientific exploration and discussion. Possibly, because the main discussants have been people interested in ICTs application in the area and so far use of ICTs, data and information in farming and agriculture was nascent.

My estimates of information costs of an agricultural product from input to consumption across the market chain is around 20-25 percent of the final cost of the product. It is more with processed foods considering packaging and advertisement.

Not understanding the economics of agricultural information constrains not only the development of agricultural information systems but also of participating in markets. For example, the cost of maintaining information for segregation, identity preservation and traceability with the commodity across the flow of market chain prevents small holders in developing countries to participate directly in export markets that essential need this information. It affects the farm input supply and access to safe foods as the cost of providing reliable quality and safety related information increases these costs. 

Most of the current “business” models of agricultural information services are based on creating a “surplus” from transforming information generated at public cost into a private good.  Private extension services take publicly available information, for example fertilizer application rates, a public good information generated by public funded research, for a crop and make it into advice for an individual farmer. The information is then specific to a particular farmer and farm and is a private good by all its characteristics. This economic model simply put is “capitalism” except that the surplus is not the classical labor but information. However, information per se, with the use of modern ICTs, is at a very low cost replicable, non-rival and non-excludable. It cannot be easily “controlled” as “property”.  Thus, there are calls for more stringent applications of intellectual property rights, copyright etc. and often masked as infringement of privacy for information and its services.

This also explains many of the statements made in the present consultation and elsewhere that 1. Agricultural business for smallholder farmers is not viable and 2. The “Government” or public sector should bear the cost for the infrastructure and other sub-structures needed for providing these information services. 

Following this up logically, the current models of information services generally can, as Yuval Harari opined at Davos recently, lead to digital dictatorships and a dystopia described in Sci-fi novels and movies as they ask for ever increasing control of data and information. (See https://www.youtube.com/watch?v=hL9uk4hKyg4&feature=youtu.be

We need a new form of “ Democratic Socialism” if more fair, just and equitable data and information services are desired for all and more so for smallholder farmers simply because almost all information is public “property”. Again, resorting to Sci-fi, this would be a “Star Trek” society where there is no “money” and the society/community bears all costs of living.

Coming back to today’s reality, the business model proposed for the Geo-Agri Gujarat Project mentioned elsewhere is the public-private-community (P-P-C) model. In this project is the Gujarat Government bears the core costs of the information system i.e. the infrastructure and substructures needed such as the “Platform” and “Trust Center”. The rationale is that though the project aims to offer a single window for all services a small farmer needs for farming, ultimately for the underlying dataset the Government is also the largest user of the information that the project generates for governance of agriculture and rural development. And, anyway, it is already spending to collect most of this information. The GeoAgri project systemizes this data and information process and creates value not only through the more efficient process but also through allowing interoperability, sharing and exchange and analyses of “big data” for information used in managing natural resources such as water for irrigation, electricity use, farm inputs and harvest forecasts, financial services and insurance and stability of markets etc. The users of information from Geo-Agri project are many in the public, private (fertilizer, pesticide supplier, finance and insurance, market intermediaries, storage, transport, processors etc.) sector and the community (Religious, Development NGOs etc) sectors. They subscribe for access to data. Telecommunications organizations also indirectly subsidize the GeoAgri project as the small holder farmers get information at less than the “market” price which are partially recovered from their subscribed use of the connectivity.  As already described elsewhere in the consultation, the private sector profits from providing value added information services. The Government also benefits beyond the use of information through a much more efficient Agri-food system that increases GDP, reduces food inflation and creates an economic surplus, increased revenue from taxation and reduced subsidization etc.  

Ajit Maru
Ajit MaruIndependent ConsultantIndia

We have reached almost the mid-point of this consultation. Valeria has summed up the discussion very competently.

I accept that data driven agriculture and Agri-food systems is not possible without cooperation and collaboration of all actors in and stakeholders of these systems. However, we should not club cooperation with “no competition”. We should be very careful and be specific about “no competition”.  Let me explain, all data and data sets collected and collated are notional based on a framework developed by individuals or a group, either with a mandate or without. If we state that there should not be any competition, does this mean that we should not have any change in this framework of data collection? The consequence of “no competition” would mean freezing the framework and with it the data and dataset. This is akin to saying that “Standards” should not change and are set in stone. This, in turn, will block innovation and possible efficiencies emerging from innovations.  Farming, agriculture and Agri-food systems are constantly evolving and need continuous change and innovation in data and applications as new needs arise. How then will we cope with this change?

In my opinion, the most efficient way would be to develop platforms that “point” to A. The applications and B. The data needed by these applications for use by the various actors and stakeholders to an Agri-food system.  This allows innovation of all three, the data, the applications and the platform itself.

The Platform would be a “Trust Center” for the meta data to individual data providers as also for applications using the data. It will assure trustworthiness, security and, where needed, privacy for both the data and application. The Platform will also function as an “App Store”, something like Apple’s App store, where users can buy both the data and the applications they need. The platform will also prescribe the standards for interoperability, sharing and exchange of data between the applications.  These functions of the platform will thus provide an open market for both data and applications and therefore force competitiveness for the benefit of the users who will also get the flexibility through mix and match of interoperable data and applications so necessary in farming and the larger Agri-food system. The platform would be mandated in its functions by the Government. For small holder farmers, who may not be able to afford the use of data and applications or the information services, the Government and other agencies may offer support or subsidies.

Let me explain this further by the “CIARD.RING” led by GFAR. CIARD.RING has been envisaged as a “Trust Center” for agricultural research for development (ARD) related data and information which includes documents as also data sets. It collects and organizes meta-data about data and information repositories not the primary data which is maintained by the repositories themselves. It provides many of the services mentioned above. CIARD.RING encourages development of applications to use ARD information and data.

The “GeoAgri Gujarat” Project, now in development in Gujarat and where I am involved, is based on this concept of a platform as described above. It aims to provide a single window access to all services farmers (overwhelmingly smallholders) in Gujarat need related to their farming. Now, the Government is funding the creation of this platform. The initial data set is being defined by a Project Steering Committee and the applications needed initially are being developed and implemented by the private sector as the GeoAgri Gujarat is expected to be financially self sustaining and is a public-private-community partnership. Initially the data used by GeoAgri is expected to be “owned” by the Government and managed by the Project Steering Committee using private sector services.

Ajit Maru
Ajit MaruIndependent ConsultantIndia

From the discussions so far, in my opinion, we are percolating to a viewpoint that smallhoder farmers will need to cooperate/collaborate to practice and benefit from data driven agriculture. I would go a step further and state that the smallholders need to aggregate in all their functions related to farming. We are all aware that Soviet style collectivisation failed. But there were many aspects of it, for example pooling of all farm resources and division of labour, that were useful and would have been successful if they then had access to the new technologies we now have for data driven agriculture. 

If we use available technologies, we can "virtually" aggregate smallholder farmers into larger producing units without disturbing ownership of land as was done in the Soviet collectives. Virtual aggregation would mean larger scale planning of farmland and cropping cycles, using technologies such as predictive planting for both the most suitable crop and the most suitable crop variety for a particular plot and field. With large scale planning, access and use of farm inputs would be more economical and efficient and outputs, and because of more precise forecast for quantum and schedule and assurance of quality of the farm product, it could be better marketed. In cultivation, the crops can be better monitored and human drudgery eliminated through use of automatic and autonomous farm machinery. These are all the issues smallholder farmers want solved.

What would such a scenario require? At the policy level, advocacy and incentives for farmers to virtually aggregate,  subsidies for virtually aggregate farmer/producer companies for start up and use new technologies etc. We would not have to worry about the current frame of land ownership being parallel to data ownership as virtually aggregation would also mean data and information aggregation. Virtual aggregation will contribute to reinforce most societal ethics of equity, fairness and justice and as also the right of smallholder farmers to live with dignity. Even many of our civil laws would remain unchanged.

Ajit Maru
Ajit MaruIndependent ConsultantIndia

Why are farmers not consulted and included in decision making that impacts them?

In my opinion, this is the outcome of prevailing economic systems that have moulded the social structures and mores for farmers in most communities. For example, laws for inheritance of land in agrarian societies are aimed more at disaagregating farmers even at the family level than aggregating them. There were exceptions, for example, in the Inca system where land was owned by the state, the religious organisation and the community  (not individually). Labour was the capital contributed to the society which was econmically based around trade and sustainable, subsistence based community farming. Apparently the entire community presided by the head of the community (or family) was involved through contributing labour in the decision making. We see similar structures and social mores in pastoral societies.

In India, where at one time when a village was considered an economic, self sustained entity, there was a Panchayati system where decisions were made by the community lead by more or less five leaders (hence the term "panchayat"; panch meaning five). Today, while this system exists, its real value in decison making about the village has become facile. 

When agriculture becomes market oriented and markets are globalised, decision making becomes complex. The farmer is technically only one actor in this system. Practically, the consumer is the most important entity making decisions for the market. Let us look at GAP (Good Agricultural Practices) certification required for participation in European agricultural commodities markets. It is the consumer who decides what are good agricultural practices and what information is needed to be assured that the practices they demand are provided. Either the farmer agrees to provide this information or is denied participation. There is no need from the consumers perspective to negotiate the information needed from farmers. This is a power based relationship. If one farmer or even farmer group does not provide the data, the consumer goes to another alternate  producer who provides the data and information.

While this does not mean I agree that farmers should not be included, I am just pointing out what in my opinion are the reasons for their non-inclusion. Market driven agriculture and globalised agricultural market diminish the role of an individual farmer to make decisions on what, how and when he/she needs to provide data and how it will be used. If we want farmers to be included, we will need to change marketing structures. For example, if policy promotes local farmers markets rather than supermarkets, farmers automatically get involved in decison making about data and information they want or need to share.