Scope of AI in Agriculture
Agriculture is seeing rapid adoption of Artificial Intelligence (AI) and Machine Learning (ML) both in terms of agricultural products and in-field farming techniques. Cognitive computing, in particular, is all set to become the most disruptive technology in agriculture services as it can understand, learn, and respond to different situations (based on learning) to increase efficiency. Providing some of these solutions as a service like a chatbot or other conversational platform for the farmers will help them keep pace with technological advancements as well as apply the same in their daily farming to reap the benefits of this service. Harnessing the power of Artificial Intelligence in agriculture will require the integration of diverse AI tools across the entirety of the agricultural production system and the adoption of enhanced systems-based approaches to ensure it is scalable, adaptive, and sustainable. AI can further accelerate the process of creating region-specific crops and livestock by harnessing the tools of machine learning.
The phrase “Right Place, Right Time, Right Product” sums up precision farming. This is a more accurate and controlled technique that replaces the repetitive and labor-intensive part of farming. It also provides guidance about crop rotation, Key technologies that enable precision farming are given below:
- The high precision positioning system
- Automated steering system
- Sensor and remote sensing
- Integrated electronic communication
- Variable-rate technology
Goals for precision farming:
- Profitability: Identifying crops and market strategies as well as predicting ROI based on cost and margin.
- Efficiency: By investing in precision algorithms, better, faster, and cheaper farming opportunities can be utilized. This enables overall accuracy and efficient use of the resource
- Sustainability: Improved social, environmental, and economic performance ensures incremental improvements each season for all the performance indicators
Use of Drone in Agriculture
Drone-based solutions in agriculture have a lot of significance in terms of managing adverse weather conditions, productivity gains, precision farming, and yield management. Before the crop cycle, drones can be used to produce a 3-D field map of detailed terrain, drainage, soil viability, and irrigation. Nitrogen-level management can also be done by drone solutions. Aerial spraying of pods with seeds and plant nutrients into the soil provides necessary supplements for plants. Apart from that, Drones can be programmed to spray liquids by modulating distance from the ground depending on the terrain. Crop Monitoring and Health assessment remain one of the most significant areas in agriculture to provide drone-based solutions in collaboration with Artificial Intelligence and computer vision technology. High-resolution cameras in drones collect precision field images which can be passed through convolution neural networks to identify areas with weeds, which crops need water, plant stress level in the mid growth stage. In terms of infected plants, by scanning crops in both RGB and near-infrared light, it is possible to generate multispectral images using drone devices. With this, it is possible to specify which plants have been infected including their location in a vast field to apply remedies, instantly. The multispectral images combine hyperspectral images with 3D scanning. techniques to define the spatial information system that is used for acres of land.
Yield Management Using AI
The emergence of new-age technologies like Artificial Intelligence (AI), Cloud Machine Learning, Satellite Imagery, and advanced analytics are creating an ecosystem for smart farming. The fusion of all this technology is enabling farmers to achieve higher average yields and better price control. Microsoft is currently working with farmers from Andhra Pradesh to provide advisory services using Cortana Intelligence Suite including Machine Learning and Power BI. The pilot project uses an AI sowing app to recommend sowing date, land preparation, soil test-based fertilization, farmyard manure application, seed treatment, optimum sowing depth, and more to farmers which has resulted in a 30% increase in average crop yield per hectare. Technology can also be used to identify optimal sowing period, historic climate data, real-time Moisture Adequacy Data (MAI) from daily rainfall and soil moisture to build predictability and provide inputs to farmers on ideal sowing time.
Use of Robotics in Agriculture
Robotics and Autonomous Systems for Livestock
At the farm level, robotic systems are now commonly deployed for milking animals. The take-up is a relatively small percentage at the moment, but an EU foresight study predicts that around 50% of all European herds will be milked by robots by 2025. Robotic systems are starting to perform tasks around the farm, such as removing waste from animal cubicle pens, carrying and moving feedstuffs, etc. Systems are in use and under development for autonomously monitoring livestock and collecting field data, all commercially useful for efficient and productive livestock farming. There are further opportunities to apply more advanced sensor technologies, combined with more autonomous systems, to perform tasks on the farm. This applies to both extensive production and intensive (indoor) systems.
A further application for robotic systems concerns the management of farmed animals, such as dairy cattle, pigs, and chickens, where intervention via the provision of appropriate and timely data can help reduce waste and environmental pollution as well as improve animal welfare and productivity on the farm.
Robotics and Autonomous Systems for Aquaculture
Aquaculture production is already a vertically integrated and professional supply chain but operates in an environment with a number of challenges that limit production, where sensors and robotic systems can play a role. Any systems deployed are naturally required to be more robust to extreme conditions and environments. The environment for aquaculture is often hostile and difficult to access by human operators with remote locations and inclement weather, with access only by small boats, leading to high operating costs and significant health and safety issues. The use of autonomous sensing and remote operation could significantly reduce the requirement for an on-site human presence making such facilities safer and easier to manage. Major challenges include environmental and health issues, such as algal blooms, sea lice, and gill diseases.
Big Data in Agriculture
Agricultural Big Data are known to be highly heterogeneous. The heterogeneity of data concerns for example the subject of the data collected (i.e., what is the data about) and the ways in which data are generated. Data collected from the field or the farm include information on planting, spraying, materials, yields, in-season imagery, soil types, weather, and other practices. There are in general three categories of data generation: (i) process-mediated (PM), (ii) machine-generated (MG), and (iii) human-sourced (HS).
PM data, or the traditional business data, result from agricultural processes that record and monitor business events of interest, such as purchasing inputs, feeding, seeding, applying fertilizer, taking an order, etc. MG data are derived from the vast increasing number of sensors and smart machines used to measure and record farming processes; this development is currently boosted by what is called the Internet of Things (IoT). MG data range from simple sensor records to complex computer logs and are typically well-structured. HM data is the record of human experiences, previously recorded in books and works of art, and later in photographs, audio, and video. Human-sourced information is now almost entirely digitized and stored everywhere from personal computers to social networks. HM data are usually loosely structured and often ungoverned.
Challenges of AI Adoption in Agriculture
Though Artificial Intelligence offers vast opportunities for application in agriculture, there still exists a lack of familiarity with high tech machine learning solutions in farms across most parts of the world. Exposure of farming to external factors like weather conditions, soil conditions, and the presence of pests is quite a lot. So, what might look like a good solution while planning during the start of harvesting, may not be an optimal one because of changes in external parameters. Artificial Intelligence systems also need a lot of data to train machines and to make precise predictions. In the case of vast agricultural land, though spatial data can be gathered easily, temporal data is hard to get. For example, most of the crop-specific data can be obtained only once in a year when the crops are growing. Since the data infrastructure takes time to mature, it requires a significant amount of time to build a robust machine learning model.
The future of farming depends largely on the adoption of cognitive solutions. While large scale research is still in progress and some applications are already available in the market, the industry is still highly underserved. When it comes to handling realistic challenges faced by farmers and using autonomous decision making and predictive solutions to solve them, farming is still at a nascent stage. In order to explore the enormous scope of Artificial Intelligence in agriculture, applications need to be more robust. Only then will it be able to handle frequent changes in external conditions, facilitate real-time decision making, and make use of appropriate framework/platform for collecting contextual data in an efficient manner. Another important aspect is the exorbitant cost of different cognitive solutions available in the market for farming. The solutions need to become more affordable to ensure the technology reaches the masses. An open-source platform would make the solutions more affordable, resulting in rapid adoption and higher penetration among the farmers.