Sustainable Agriculture Leveraging Artificial Intelligence Systems in Kenya's Agri-food Supply Chain

The Agro-food supply chain is crucial for achieving Sustainable Development Goal No. 2 of zero hunger and sustainable agriculture. However, Kenya faces significant post-harvest losses, mainly attributed to challenges in first and last-mile logistics. In the era of technological advancements, this research paper explores the potential of Artificial Intelligence (AI) to enhance the Kenyan agri-food supply chain. Building on existing information, the study focuses on AI's role in monitoring and controlling farmland outputs, optimizing supply chain logistics, and addressing fraud and counterfeiting. The research objectives include assessing AI's utility in monitoring and controlling outputs in farmlands improving supply chain efficiency, and combating fraud in agricultural inputs. Research methods involve a comprehensive literature review, analyzing case studies such as Project FARM and FAO's Fall Armyworm Monitoring and Early Warning System, and reviewing scholarly articles on AI applications in agriculture. The research results highlight the benefits of leveraging AI in farmland monitoring, climate change adaptation, supply chain logistics, and fraud prevention. AI technologies can enhance agricultural productivity, reduce transportation costs, and eliminate corruption in the supply chain. The findings suggest that integrating AI systems into the agri-food supply chain is vital for achieving sustainable agriculture in Kenya. The study concludes that AI offers innovative solutions to address the challenges faced by smallholder farmers, enhance supply chain efficiency, and contribute to achieving zero hunger and sustainable agricultural practices.


INTRODUCTION
Sustainable agriculture has become a critical component in addressing global food security and environmental concerns.In the context of Kenya, where agriculture plays a significant role in the economy, the integration of artificial intelligence (AI) systems holds promise for enhancing efficiency, productivity, and environmental sustainability in the agri-food supply chain (Kipkogei et al., 2021).Kenya's agricultural sector is characterized by smallholder farmers who face challenges such as limited access to modern technology, fluctuating weather patterns, and inefficient supply chain management (World Bank, 2019).These challenges contribute to food insecurity and hinder the sector's overall sustainability.Therefore, the convergence of sustainable agriculture and artificial intelligence holds great potential for transforming Kenya's agri-food supply chain.By addressing current challenges and leveraging AI technologies, the country can requiring human intervention.
Source: https://www.abetasquare.com/understanding-artificial-intelligence/ The Role of Artificial Intelligence in Agri-food supply chain Artificial Intelligence (AI) plays a pivotal role in revolutionizing the agricultural sector, contributing to the development and implementation of sustainable practices.AI technologies offer innovative solutions for addressing various challenges in agriculture, ranging from resource optimization to crop management.The global population is projected to increase from 7.7 billion to 9.7 billion by 2050, as reported by the United Nations.Simultaneously, the ongoing impact of global warming is anticipated to lead to a scarcity in agricultural production in the coming decades.
Addressing this challenge requires leveraging the advancements in digital technologies, with a particular focus on incorporating Artificial Intelligence (AI) solutions into the agricultural supply chain.(FAO,2018).Industry 4.0 represents a technological endeavor focused on transforming manufacturing by digitally reshaping both processes and products.This initiative anticipates a comprehensive evolution of the industrial supply chain, envisioning increased autonomy and intelligence throughout (Liu et al,2021).Agri-food supply chains are encountering comparable challenges in embracing technological advancements, including the integration of the Internet of Things (IoT), robotics, artificial intelligence (AI), big data and analytics, and blockchain.Nevertheless, within the realm of essential cross-industry technologies, artificial intelligence (AI) is just starting to surface as a viable solution to address various challenges encountered by the agricultural sector in

AI and the monitoring and control of outputs of farmlands
Although agriculture is the backbone of Kenya's economy, agricultural productivity stagnates due to a lack of proper monitoring and control of production.Besides, the country lacks the required technology to predict climate change and provide soil and crop dynamics knowledge.
Considering that most farmers in Kenya are smallholder farmers between 5 million and 9 million (Wight, 2019), the agri-food supply chain is highly vulnerable to climate change and soil and crop dynamics.Consequently, digital technology such as AI could offer an innovative approach to solving such uncertainty in climate change and crop and soil dynamics.Artificial Intelligence (AI) plays a crucial role in the monitoring and control of outputs in farmlands, contributing to precision agriculture and sustainable farming practices.This integration of AI in agriculture involves the use of advanced technologies such as sensors, drones, machine learning, and data analytics to optimize various farming processes.Evidence mainly from developed nations supports the continuing importance of AI in the agricultural supply chain, as seen in Table 1 in the following reviewed studies.

AI and supply chain logistics
Logistics remains a central term in the supply chain.According to the Council of Supply Chain Management Professionals (CSCMP), logistics is a domain of the supply chain process.It plans, implements, and controls the forward and reverse flow and shortage of goods, services, and the related flow of information efficiently and effectively between the source point and the point of consumption to meet the customer's needs (CSCMP, 2018).Meanwhile, professors of Michigan State University define logistics as activities such as packaging, transportation, warehousing, and others that move and position inventory while acknowledging its role in supply chain synchronization.
vehicles, sparse distribution of farmers, and poor storage options remain major challenges raising per unit cost at every stage of the chain.
Despite the first and last-mile logistics being core to rural supply chains and crucial for smallholder livelihoods, they remain inefficient and expensive.For instance, research shows that approximately 62 percent of farmers use manual forms of transport, thereby consuming more time (AFCAP, 2014).Moreover, the Rural Transport and Agriculture Fact Sheet, 2015 documents that transportation costs make up 28% of Kenya's final market prices, significant in the value chain (Olwande et al., 2015).How then can AI be leveraged to lead to notable savings in the Agri-supply chain?
Research documents evidence of the importance of AI in supply chain logistics as demonstrated in the following reviewed studies in

AI and Agri-food supply chain Fraud and Counterfeiting
Fraud and counterfeiting are major challenges in the agri-food supply chain in Kenya.
Right from counterfeit Agro-inputs such as seeds and fertilizer to fraudulent pesticides, the chain is bedeviled by fraudulent activities at all stages.Although quality agricultural inputs, including fertilizer, seeds, and pesticides, are critical to agricultural productivity, access to high-quality inputs remains challenging in rural African markets.For instance, a previous study in Kenya determined that only close to 77 percent of hybrid maize seeds germinated or grew into maize crops, even though the Kenyan government requires a minimum of 90 percent for certification (Miguel, Hsu & Wambugu, 2020).The question then is how to leverage systems that eliminate fraud and counterfeiting in Kenya's agri-food supply chain.A large body of research demonstrates in Table 4 the viability of AI to act as an anti-corruption tool in developed settings.Note: The table summarizes the authors, context, and utility of artificial intelligence in various agricultural domains, emphasizing its role in fraud and counterfeit detection.

RESULTS AND DISCUSSION
From the two sets of reviewed studies, it is apparent that the Kenyan agri-food supply chain can benefit from leveraging artificial intelligence to monitor and control food production.Indeed, evidence in the Kenyan context documents the potential inherent in AI in monitoring and controlling food production.For instance, project FARM (Financial and Agricultural Recommendation Models) is a project being undertaken in the Kakamega region under the collaboration of Capgemini, a French consultancy firm, and Agric, an East African social enterprise to advise farmers on appropriate planting times (Wight, 2019).In this project, Capgemini uses AI to crunch farming data and send insights to farmers' cell phones.Through their testimonies, farmers participating in the project are upbeat about the benefits.They receive an SMS on how to tend their farm work and farm projects.
Another example of the potential of AI is FAO's adoption of the technology to fight the Fall Armyworm (FAW) spread that has recently plagued East Africa (Bramhall, 2021).FAO developed the Fall Armyworm Monitoring and Early Warning System (FAMEWS) mobile app, which tracks the insect's changes over space and time to gain knowledge of its behavior in a new context and inform the best response.This way, farmers and agricultural extension workers directly manage their crops to prevent further infestations and reduce damage.
With such evidence on the potential of AI to provide necessary information for farmland evaluation and control, we argue that sustainable agricultural development in Kenya requires research to focus on Agri-tech systems such as AI, which can revolutionize the agri-food supply chain.Lack of farmland monitoring and control has resulted in poor soil fertility associated with low yield and land degradation (Birch, 2018).The absence or underuse of inorganic and organic fertilizers, continuous cropping, and soil erosion are associated with less than two tons per hectare of maize yield among smallholders in Kisii compared to the required nine tons (Mulinge et al., 2016).Therefore, AI is poised to control soil fertility by using modern soil testing methods that calculate parameters such as temperature, pH level, moisture content, organic matter, NPK, unpredictable temperatures that threaten agricultural productivity.Therefore, leveraging AI is a sure way of forecasting and predicting weather patterns.Research shows that the nowcasting AI system can expect more accurate short-term rain predictions, including storms and floods (Bochenek & Ustrnul, 2022).The predictive analytics advanced by AI enables farmers to acquire information on the timing of sowing the seed to maximize yield.Pricing of harvested maize remains a contentious issue in Kenya.Such instability in pricing has not allowed farmers to plan their production.Yet AI can guide farmers on the demand level, future price patterns, and crop type to sow for maximum rewards.
From the above research evidence, we posit that sustainable agricultural production systems in Kenya require that the logistical challenges farmers experience, especially in the first and last-mile logistics, be eliminated.While this may prove difficult, we argue that technology, especially AI and the Internet of Things (IoT), can be leveraged in the logistics of the agriculture supply chain.
For instance, AI can handle mass data, making it highly effective in inventory management.It is further argued that intelligent systems can expedite the analysis and interpretation of huge data sets, enabling timely demand and supply forecasts (Preil & Krapp, 2022).Through their intelligent algorithms, AI systems facilitate the prediction and discovery of new consumer habits and enhance visibility and transparency into all aspects of the supply chain.Suffice it to say that AI could be the panacea to dirty deals that threaten Kenya's food security.
Indeed, as reported in The East African Newspaper dated October 17, 2018, various forms of corruption undermine Kenya's food security (Mukami, 2018).For instance, in 2018, the National Cereals and Produce Board (NCPB) had some of its officials colluding with traders disguised as farmers to embezzle Ksh1.9 billion.In another incident, a South African company was awarded a multi-million shilling deal by senior state officials to supply Kenya with maize.
However, the award process was fraught with irregularities (Mukami, 2018).Such dirty deals could be eliminated by leveraging AI to enhance supply chain visibility.With it, stakeholders can track produce as it travels from supplier to manufacturer or NCPB in this case, and then to the consumer.
Farmers in Kenya have often spent endless days and nights in queues trying to deliver their produce to the NCPB and they spend colossal amounts of money.However, such inefficiency in farmlands should give further incentives for leveraging AI systems.The ability of AI to forecast food shortages places this technology at the forefront of ending hunger, improving nutrition, achieving food security, and promoting a sustainable, agro-food supply chain.Moreover, AI systems provide data that can be used to analyze scenarios efficiently, anticipate potential risks and take remedial action to improve the agri-food supply chain.
Another major contribution of using AI systems in the agri-food supply chains is efficiency in supply chain logistics.Ending hunger means having food reach everyone.Therefore, AI systems can guarantee that foods reach everyone by navigating the challenges farmers experience in the first and last-mile logistics.Produce, food transportation and tracking appear to improve when AI-controlled drones or other transportation modes automate service delivery.
Moreover, such an efficient transportation network alleviates inventory predictions by enabling ease of goods tracking.It is argued that suppliers and consumers can track food or produce sources to identify organic and non-organic products (Gui-e & Jian-Guo, 2020).Still, on the agri-food supply chain, AI can maximize delivery routes, minimize fuel expenses and ensure quick delivery times.Therefore, artificial intelligence systems are likely to lower costs experienced with intermediaries in the first and last-mile logistics.Meanwhile, AI optimizes warehousing and storage.Through the genetic algorithm (GA) and radial basis function (RBF), AI can develop time series forecasting models for perishable produce (Niu & Feng, 2021).
The other elements AI systems are bound to impact in Kenya are fraud and counterfeiting in agricultural inputs.Numerous cases of food fraud and counterfeit inputs can be stopped through AI.It is refreshing to realize that efforts are on in Kenya to address fake inputs through mobile technology, albeit minimally.These efforts demonstrate the desire to improve the agri-food supply chain through AI systems.Although we acknowledge that it may ultimately be difficult to comprehensively integrate machine intelligence in the agri-food supply chain, we posit that sustainable agriculture is achievable by engaging machine intelligence to cut down costs in the first and last-mile logistics and maximize farming yield.
development and distribution of precision farming systems capable of supporting decisions companies make regarding risk and disease prediction enhanced transparency, enhanced last-mile delivery, personalized solutions to upstream and downstream stakeholders, and facilitates an agile procurement strategyPandian (2019) Warehousing efficiency Enhancing logistics, coordination, and management potentials, i.e., creating a smart warehousing environment.scarcity in high-skilled personnel -AI allows for automated truck driving in logistics which is important for the design of autonomous driving supervision Han as AI enable keeping the required quality and quantity and sequencing and scheduling profiles.Moreover, fully automated processes secure transportation and safeguard workers from the monotonous repetition of activities.Semi-autonomous transport vehicles can be used such that a control system informs the driver of the most optimal route.vision systems (CVSs) allows a non-destructive evaluation of quality levels in vegetables.Has the capability to monitor quality level regardless of packaging.

A
report titled Counterfeiting in African Agriculture Inputs-challenges & Solutions (de Boef et al., 2019) paints a grim picture of counterfeit inputs in rural African markets.According to CropLife Middle East Africa (2011), 30 percent of pesticides on sale in Ghana by 2011 were unlicensed or smuggled.On the other hand, the Kenya Agricultural Research Institute (2012) indicated that as of 2012, there were 40 percent fake seed packets in Kenya, and by 2014 there were 30 percent counterfeit hybrid high-yielding variety seeds in the Ugandan market.According to Kenya's Anti-Counterfeit Authority (ACA), there has been an upsurge in reported cases of counterfeit targeting the grain-basket counties.Such cases involve counterfeit fertilizers and seeds, fraudulent labeling, trademark infringements, and theft of labels and packaging materials (ACA, 2021).
such as Deep Belief Networks (DBN) enable the construction of robust computer vision for precision Agriculture Musra et al. (2020) Agriculture and food stability AI alongside IoT and big data enhances greenhouse monitoring, drone-based crop imaging, food quality assessment, and food safety Jim et al. (2020) Vegetable farming The use of AI increases the accuracy of vegetable monitoring of fertilizing and sowing.Which allows for the detection of General Agriculture AI addresses disease management, counterfeit inputs, crop growth management, and soil properties.

Table 1 .
Importance of artificial intelligence in monitoring and control of outputs in farmlands

Table 2 .
Importance of Artificial intelligence The table summarizes the importance of Artificial Intelligence in agriculture, highlighting specific contexts and the utility of AI as reported by various authors.

Table 3 Table 3 .
Importance of Artificial Intelligence in Supply Chain Logistics

Table 4 .
Artificial Intelligence's Role in Fraud and Counterfeit