Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Problem Space

Why are we doing this?

Problem statement:

How we may enable farmers to assert control on their data?

Impact of this problem:

Farmers, specially small holder in developing countries have to make multiple decisions that affects their livelihood directly. With agriculture tech ecosystem getting more focus, the importance of data is increasingly becoming important in aiding the decision making process. The data typically involves sensitive information and also transaction information which can’t be shared without consent from the farmers. Sharing this information unlocks potential of new services for the farmers as well as ecosystem actors as they can bundle services together leading to better experience of the farmers.

How do we judge success?

  1. X number of farmers consent to share their information maintained by one org to another to get a service

  2. Some percentage of farmers avail the service

  3. Some percentage of farmers get benefit from the service

  4. Org maintaining data and providing service see value with improved farmer satisfaction

What are possible solutions?

There are two broad ways to solve this problem:

  1. Each application provider publishes the data back to the farmers who maintain the data in some data wallet and share it at their own convenience

  2. Today most of the applications ask permissions to capture information t of the farmers at the time of onboarding. The farmers, in the same way can give consent to share the data captured to avail some services beyond what is provided by the application. The consent can be a token that is collected against the user. The data itself is not maintained by the farmers but in stead they control with whom and for what it is shared.

Validation

What do we already know?

  1. Globally, GDPR is being seen as a reference document and GDPR focuses heavily on consent and restricting purpose, time, storage etc.

  2. Government of India has already come up with a consent framework, here are some resources:

    1. Presentation: https://www.slideshare.net/ProductNation/data-empowerment-protection-architecture-depa

    2. They are already working on integrating a confidential compute with the consent framework

    3. Agriculture and Healthcare are the focus area with adoption in banking sector already setting precedence

    4. A pilot is being proposed in the state of Karnataka that takes consents from the farmers

  3. There are group of startups who want to exchange data based on the consents

What do we need to answer?

  1. How does Can consent relate be extended to usage policies? Right now, the consent typically involves sharing data and once data is shared it is lost. On the other hand usage policies restrict the use of data while sharing. Come up with a use case that demonstrates the consent driven usage policies. Define clear value add for the farmers and orgs who want to use this.create usage control (consent is used to transfer data but can not enforce how, to whom and for what data is to be transferred)?

  2. How do the farmers and/or organisations get value if there is a way to enforce “purpose” in the consent artefact?

  3. How do we get informed consent?

  4. How do we raise awareness for enforcing usage control on data?

  5. Is the usage control imposed on provider or consumer or both (provider is collecting/ managing data but the true owner is the farmer)?

Ready to make it

What are we doing?

We have two use cases running in India. One in the state of Bihar and another in the state of Andhra Pradesh.

  1. In use case 1, Digital Green provides coaching to farmers on climate smart practice. Digital Green already maintains database of farmers, the practices they have adopted and the crops they have grown. While showing the videos, the field extension worker gets the consent from the farmers to share their activity and field details to certifiers and buyers. If the farmers give the consent, the relevant data is made visible.transferred.

    1. Incentive for farmers: ease of onboarding and benefit of getting discovered

    2. Incentive for organisations: buyers get to discover the farmers easily

    3. incentive for govt: give control in the hands of farmers

  2. In use case 2, Digital Green right now is operating the certification shops for chilli farmers where the grade of produce grade is quantified and they get customised advisory to improve the quality and also to connect to the buyers who can give better price. In future this is to be operated by government owned shops where the details of the grade will be shared based on the consent of the farmers.

    1. incentive for farmers: decide whom and what data they want to share, if grade is bad they want to share to just the advisory service provider details about the practice

    2. incentive for the buyers: discovery of farmers and able to offer better price for better quality

    3. incentive for the govt: enable network and help improve quality in long term

Why will a customer want this?

  1. Use case 1:

    1. Farmers don’t want to share details about their field size etcBuyers need the outcome - a score of how their activity has impacted environment and not the field size and entire details so that they can offer prices to manage carbon creditsare sensitive about sharing the details of their field size etc unless there is benefit

    2. Government who is maintaining this data with digital green wants to impose necessary restrictions so that farmers data is controlled by the farmers

  2. Use case 2:

    1. Farmers who get lower grade may not want to share the data to the buyers as it gets lower price but they may want to share the details with an advisory service to know how they can improve their grades

    2. The grade of the produce by the farmers is a sensitive information that needs to be managed by the government and they want necessary restrictions to be placed

Visualize the solution

  1. Use case 1

    1. The data of extension workers who show videos to the farmers about their farm and farming activity (including history) is shared through trusted connector which runs the application to give the output of Greenhouse gas emission score against the farmer

    2. The score of the farmers can be shared to a match making platform where buyers can contact farmers with good score

    3. Consent manager is like a service that can be embedded in any application (like a payment service)

      1. while showing videos to the farmers on benefits of climate smart practice, the field worker can mention interest.

      2. While the data about the farmers who have seen video is entered, the consent manager asks for consent (using SMS/IVR + application) from farmers if they want to share their information to share their score to interested buyers.

      3. The information shared for calculating score and shared to the buyers is told. Consent manager will have some interface to inform and educate farmers.

      4. If the farmers give consent, a token is generate that defines the consent to share data to be used by an application.

      5. The receipt of token triggers data sharing through the connector with a usage policy that allows processing data by the scoring application and if there is a valid token.

    Use case 2

    1. When the farmers get their produce verified they give consent to share what parameters they want to share for receiving advisory services to improve their quality and get better price.

    2. The data for the farmers giving consent is shared through a connector that restricts usage of data by match making application, that is, matches content provider and buyer to the farmers and sends them a notificationand provide advisories collect consent using an app:

      1. The app has visual images or videos that explain the data that is held by the govt, how sharing this can help the farmer and the reason for their consent

      2. Explain that the farmers will get verification call and/or SMS and the response they should give

      3. Stage 1 focuses on enabling farmers to exercise control of “with whom and how data can be shared”

      4. Stage 2 enables farmers to exercise control of “what part of data can be shared to whom”

      5. Stage 3 enables farmers to exercise control of “the purpose of sharing specific part of data - defined by application”

    3. The data of the farmers about their farm and farming activity (including history) is maintained by DG in MySQL DB and an application is created that queries data from the MySQL tables to create a combined table of relevant columns. The application runs inside a connector and passes on only that field where the consent artefact is a) obtained and is b) valid. The application need not be real time and can have some latency in fetching data

    4. Every time data is shared for a specific farmer, the transaction is logged which has the details of a) the connector or endpoint where it was made available, b) the metadata of the org who had the access.

    5. The connector can be run by DG or by the government partner.

  2. Use case 2

    1. Similar to above except that the farmers give consent at the certification shops.

Scale and effort

  1. Use case 1:

    1. Scale: currently in very initial scoping phasewe wish to start in a toned down use case where we start with advisory creation by sharing location data for getting soil parameters (start with few 100 farmers)

    2. Effort: highMedium

  2. Use case 2:

    1. Scale: pilot done with 2000 farmers

    2. Effort: mediumMedium - large

Additional resources

Depa book:

...

Detail presentation of use case here:

...

...