Turbocharging Digital Transformation in Emerging Markets: Unleashing the Power of AI in Agritech (ITC)

4 Dec 2023 13:00h - 14:00h UTC

official event page

Table of contents

Disclaimer: This is not an official record of the UNCTAD eWeek session. The DiploAI system automatically generates these resources from the audiovisual recording. Resources are presented in their original format, as provided by the AI (e.g. including any spelling mistakes). The accuracy of these resources cannot be guaranteed. The official record of the session can be found on the UNCTAD website.

Full session report

Martin Labbé

During the discussion, speakers explored the application of blockchain technology in various sectors. It was highlighted that blockchain provides a decentralised database and ensures data integrity. This technology is particularly relevant in ensuring traceability within the new European Union (EU) regulatory framework. By utilising blockchain, companies can securely store and track information, allowing for more transparent and accountable processes.

Another topic that was highlighted during the discussion was the use of PharmaConnect’s solution for shea nut collectors. This innovative solution enables offline data collection, which can be synchronised and uploaded onto the blockchain once the collectors return to their offices. By employing this solution, shea nut collectors in Ghana and other regions can efficiently collect data and ensure its accuracy, thereby supporting their livelihoods.

The speakers also emphasised the potential of various digital technologies, such as artificial intelligence (AI) and blockchain, in supporting smallholder farmers to improve their agricultural practices. For instance, AI platforms can analyse satellite imagery and weather data to provide insights on optimal planting times and crop choices, thereby aiding in agricultural sustainability and profitability. Furthermore, companies like Descartes Lab and AgroCare are successfully leveraging AI to provide services to farmers and enhance their productivity.

However, it was recognised that agri-tech startups face challenges in creating sustainable business models and generating revenue, particularly when working with smallholder farmers. These startups often rely on other stakeholders, such as mobile network operators, input providers, and banks, to fund their services, but this support is not always consistent or sustainable. To address this issue, improved connectivity and access to digital payments were suggested as potential solutions.

The importance of digital innovations for agricultural productivity was acknowledged, but obstacles to monetising these technologies were also discussed. While there have been advancements in connectivity and digital payment access, challenges in achieving sustainable monetisation persist. This indicates the need for further exploration of business models that can effectively generate revenue from digital technologies in the agricultural sector.

The speakers also discussed the need for data accuracy and context-specific information to efficiently implement AI in different parts of the world. For instance, generative AI, such as chatbots for farmers, relies on user-collected data, and ensuring the accuracy of this data is essential for its effective use. Additionally, the role of government support in funding and infrastructure for AI in agriculture was highlighted, with the example of Descartes Lab in Mexico, which receives partial support from the government.

However, the struggle to monetise AI services in emerging economies, where smallholder farming is prevalent, was acknowledged. Unlike commercial farming setups in North America, emerging economies often lack the financial capacity to pay for these technologies, posing a challenge for startups operating in these regions.

In conclusion, the discussion revolved around the potential and challenges of digital technologies, including blockchain and AI, in various aspects of agriculture. While these technologies offer promising solutions to improve agricultural practices, issues such as sustainable business models, revenue generation, data accuracy, and monetisation in emerging economies need to be addressed. Government support and the adoption of best practices across different geographies were also identified as crucial elements for the advancement and widespread implementation of digital technologies in the agricultural sector.

Benjamin Kwasi Addom

Data, artificial intelligence (AI), and new technologies have the potential to greatly benefit agriculture by assisting farmers in making informed decisions and managing their crops more effectively. For example, in Uganda, data combined with AI has been used to support farmers in determining the best times for planting and harvesting, which can significantly improve crop yields. Furthermore, the use of data can enable the implementation of crop insurance programs and accurately predict and mitigate negative events such as droughts, providing farmers with financial support when needed.

However, the success of agritech enterprises heavily relies on data quality and access. While these technologies offer significant opportunities, they need to be complemented with accurate and reliable data. Many startups in this industry are currently facing challenges due to data issues, emphasizing the importance of ensuring data quality and accessibility for the overall success of the sector.

To support agritech companies, the implementation of policies is encouraged by the Commonwealth Secretariat. The Secretariat recognises the value and potential of these enterprises and is taking a policy angle to facilitate their growth and development. Encouraging equal opportunities for all innovators, regardless of their background or size, is another key focus. The Secretariat aims to promote a level playing field and ensure that all innovators receive fair consideration and support.

Another critical aspect to consider is the management of national agricultural data. Currently, there are challenges related to data being disaggregated and duplicated, particularly within smallholder agriculture. This can hinder the effectiveness of agri-tech business models. To address this, it is suggested that countries should enhance their national agricultural data infrastructure. By managing data at the country level, a larger and more comprehensive data infrastructure can be created, allowing AI tools and technologies to operate more efficiently.

Furthermore, it is proposed that the sharing of polygon data of farms within national data infrastructure can significantly reduce costs and duplication. Currently, multiple agri-companies in the same country often map the same field multiple times, resulting in unnecessary duplication and high expenses. Storing GPS coordinates of fields in a centralised national infrastructure can streamline information sharing, making it easier and more cost-effective for all stakeholders.

It is also crucial to tackle the issue of data duplication and the lack of a single identity for farmers within national infrastructure. Often, the same farmer’s information is duplicated in different systems within the same country, causing additional burdens and unnecessary costs. The creation of a single identity for each farmer and ensuring that it is stored and protected within the national infrastructure can help address this issue.

By addressing these challenges and improving data management and accessibility within the agricultural sector, the overall business model for startups can be enhanced. This will create a more supportive environment for innovation and advancement in agritech, leading to greater benefits for farmers and the industry as a whole.

Moreover, while AI and new technologies have significant potential in agriculture, it is crucial to understand that they should not replace human involvement but rather complement it. Artificial intelligence should be seen as a tool to support and augment human decision-making and scientific knowledge, rather than completely replacing human expertise.

In conclusion, data, AI, and new technologies offer great potential in revolutionising and improving agriculture. However, to fully harness their benefits, it is important to address challenges related to data quality and accessibility. Implementing supportive policies, managing national agricultural data effectively, and reducing duplication will contribute to the success of agritech enterprises and ensure equal opportunities for all innovators. Furthermore, while AI can enhance decision-making processes, it should be utilised as a complementary tool alongside human involvement. By effectively leveraging these resources, we can create a more sustainable and efficient agricultural sector that benefits farmers, the industry, and the wider community.

Susanne Emonet

The discussion focused on the need for traceability in supply chains to ensure responsible consumption and production. It was highlighted that traceability is challenging in complex supply chains and that it cannot be ascertained by sight if child labour or deforestation was involved in a product’s production. To overcome these challenges, blockchain technology was suggested as a solution to support traceability. It was explained that blockchain provides a shared bookkeeping system in which the same unalterable data is stored. This allows the product to be traced through traders, brands, retailers, and eventually to the consumer.

A case study on the use of offline tooling for traceability by Savanna Fruits was presented, showcasing positive results in Shea nut sourcing in Ghana. It was stated that Shea nuts are collected on public grounds where control is absent, and offline tooling helps to record transactions of both Shea and payment. This system allows training on sustainable practices, foundational records for premium payments, and the data gathered is uploaded to the supply chain and used by partners.

Another application of blockchain technology was mentioned, which focuses on ensuring proper traceability in the context of a new regulatory framework in the EU. It was explained that data is collected offline and then synchronized when collectors return to the office, after which it is uploaded to the blockchain.

Efforts to avoid child labor in cocoa farming were discussed. A client in Côte d’Ivoire was mentioned, who was willing to pay significant premiums to enable cocoa farmers to send their children to school and pay workers instead. The traceability core system was linked and enriched by a connection to flip phones, enabling direct communication and payment confirmation from the farmer’s side.

The potential of digital payments was highlighted as a promising tool for transparency in supply chains. Mobile money deployments across Africa were cited as a reliable digital payment form, and it was suggested that digital payments can replace physical payments, adding automated traceability to the transactions.

The importance of interoperability was emphasized, stating that a system capable of taking data from various sources avoids doubling of effort and data. It was also mentioned that interoperability ensures supply chains remain flexible to accommodate business needs.

Financial viability was discussed in the context of transparency technology. It was mentioned that transparency is an additional cost, so it needs to be as lean as possible to be cost-effective. Additionally, it was noted that investments in sustainability should not solely be spent on technology.

The need for data protection was highlighted, particularly when dealing with significant amounts of data. It was stated that the potential knowledge gained from data collection is too significant to not be protected. The General Data Protection Regulation (GDPR) was cited as a crucial framework for ensuring data protection.

Artificial Intelligence (AI) was discussed as a tool that should add value and address issues such as sustainable agriculture and smart agriculture. AI’s ability to enhance data validation was presented as an advantage, as it can systematically check large data pools and monitor data entry patterns for significant changes.

The role of AI in supplier selection mechanisms and risk assessments was also highlighted. It was argued that AI can improve these processes by building trustworthy data.

The value of AI for traceability and data validation was recognized. It was mentioned that AI can strengthen the trustworthiness of data and can be integrated with third-party satellite data.

Lastly, the importance of technology solutions worth paying for and involving investors from later parts of the supply chain was emphasized. It was noted that most investors come from later parts of the supply chain and that expensive technology for first-mile solutions may be limited.

Overall, the summary highlights the various perspectives and solutions discussed to enhance traceability in supply chains, emphasizing the need for responsible consumption and production. It also underscores the key role of blockchain technology, offline tooling, digital payments, interoperability, financial viability, data protection, and AI in achieving transparency and sustainability.

J.Sjaak Wolfert

The analysis emphasises the importance of organising the process of digital transformation in agriculture, rather than solely relying on the technology itself. It argues that digital transformation is more about the effective organisation of the process, taking into consideration factors such as ethical, legal, and social aspects, rather than simply focusing on the technological advancements.

Furthermore, the analysis highlights the significance of viability and robustness in innovation ecosystems for successful digital transformation in agriculture. It suggests that developing lean multi-actor approaches and considering various aspects such as ethical, legal, social, and business modelling can contribute to the creation of these ecosystems.

The analysis proposes embedding projects in a larger network of digital innovation hubs as a long-term and sustainable solution for digital innovation in agriculture. This approach allows for the utilisation of state-of-the-art knowledge required for scaling up specific solutions. The Smart AgriHubs project, funded by the European Commission, is mentioned as an example of this embedded approach.

Challenges in AI farming technology are also discussed in the analysis. The business model is identified as one of the major challenges, as the market size is not always large enough, leading to fragmentation in different types of farming. Additionally, technology providers often focus on specific groups, which makes scaling up solutions difficult. Furthermore, the high cost of solutions hinders their adoption by farmers, despite their added value.

The analysis emphasises the importance of trust in AI technologies for farmers. However, due to the complex nature of AI, explaining how the technology works can be challenging, which can affect the level of trust. Furthermore, intellectual property rights can potentially conflict with the need for explainability, posing another obstacle to trust in AI.

Data management and policy are highlighted as crucial aspects of AI farming technology. The vast amount of data collected raises questions regarding its use and whether it complies with regulations. The European Commission is working on several acts to regulate data use, considering issues such as intellectual property rights and the explainability of AI algorithms.

Combining different types of funding is suggested as a means to avoid fragmentation in the innovation process. By integrating public and private funding, the innovation process can be streamlined and more efficient.

The analysis emphasises the importance of continuous development in digital innovations. By treating digital innovations as an ongoing process, rather than a one-time implementation, continuous improvements and advancements can be made.

Lastly, the analysis highlights the significance of embedding digital innovations in effective ecosystems. This involves integrating digital innovation hubs and knowledge networks into competent centres. By doing so, the exchange of knowledge, expertise, and resources can be facilitated, leading to more successful and impactful digital innovations in agriculture.

In conclusion, the analysis stresses the importance of focusing on the process rather than solely relying on technology in digital transformation in agriculture. It underscores the significance of innovation ecosystems, embedding projects in larger networks, addressing challenges in AI farming technology, data management and policy, funding integration, and continuous development of digital innovations. These insights and recommendations provide a comprehensive understanding of the key factors and considerations for successful digital transformation in agriculture.

Elisabetta Demartis

The adoption of artificial intelligence (AI) in agriculture has the potential to greatly improve access to information and advisory services, particularly in remote areas. For example, Descartes Lab, a Mexican company supported by the government, uses AI to analyse market demand through satellite imagery and weather data. Similarly, AI-powered devices like AgroCare’s Nutrient Scanner can monitor soil health and provide estimates of missing nutrients, thereby helping farmers make informed decisions about their crops.

To ensure the full efficiency of AI in agriculture, sustainability must be prioritised alongside profitability. This involves integrating intelligence that can elaborate on collected data. AI can play a crucial role in promoting sustainable and regenerative agricultural practices, which can aid in combating climate change and related challenges. By leveraging AI, farmers can make more environmentally conscious decisions regarding resource allocation, pest control, and crop cultivation.

However, it is important to emphasise the need for human oversight in the input and output of data, particularly in generative AI systems like chatbots for farmers. Human verification is necessary to ensure the accuracy, usefulness, and context-specificity of the data used in these systems. This is crucial for building trust and effectively utilising AI technology.

The business model for AI in farming can be particularly challenging, especially for smallholder farmers in emerging economies. Monetisation of AI services may be difficult for these farmers due to their limited financial capacity. To address this, business models that can reach smallholder and low-income farmers are essential. For example, the “free option” model, where a third party such as the government, a development organisation, or a telecom operator provides and pays for the solution, can be effective. Another model is “data monetisation,” where insights from user data are sold to third parties. However, transparency, data protection, and farmer data ownership must be carefully considered in such arrangements.

The sustainability of AI solutions in agriculture is also a challenge that needs to be addressed. Ensuring the continuity of AI initiatives after project funding or donor support ends is crucial for long-term effectiveness. Therefore, business models and strategies should be developed to ensure the sustainability and scalability of AI solutions in agriculture.

Engaging communities in the creation of business models may provide a potential solution to the challenges faced in AI adoption in farming. By involving farmers and allowing them to feel like contributors to the system, their interest and participation can be fostered. This can be achieved by not only seeking their information but also involving them in buying other services offered by the system. Such community engagement has the potential to address some of the challenges related to the business model and enhance the effectiveness of AI in agriculture.

Interoperability of systems, both within countries and among countries, is crucial for efficient data management in agriculture. The ability to share and exchange data is important for effective decision-making and collaboration. By ensuring interoperability, AI can be used more efficiently to drive innovation and improve agricultural practices on a global scale.

It is important to remember that the purpose of AI in agriculture should be to support human tasks, not replace them. Jobs within the agri-food value chain, such as advisory services, should be maintained to promote decent work and economic growth. AI should be seen as a tool to assist humans, enhancing their capabilities and decision-making processes.

Finally, it is crucial to develop AI solutions that can address the challenges faced by marginalized groups, including women, vulnerable individuals, and those affected by displacement, climate change effects, and conflicts. By addressing the specific needs and circumstances of these groups, AI can contribute towards achieving gender equality, climate action, and peacebuilding initiatives.

In conclusion, AI has the potential to revolutionise agriculture by providing access to information and advisory services, improving sustainability practices, and addressing the specific needs of marginalized groups. However, challenges such as business models, sustainability, and data management need to be overcome for the full realisation of AI’s benefits in agriculture. It is important to approach the adoption of AI in agriculture with careful consideration of its ethical implications and a focus on enhancing human capabilities and inclusivity.

BK

Benjamin Kwasi Addom

Speech speed

154 words per minute

Speech length

1612 words

Speech time

628 secs

ED

Elisabetta Demartis

Speech speed

122 words per minute

Speech length

1407 words

Speech time

692 secs

JW

J.Sjaak Wolfert

Speech speed

159 words per minute

Speech length

2277 words

Speech time

859 secs

ML

Martin Labbé

Speech speed

168 words per minute

Speech length

1762 words

Speech time

630 secs

SE

Susanne Emonet

Speech speed

174 words per minute

Speech length

2092 words

Speech time

720 secs