Farmers face hefty challenges to AI adoption in agriculture sector
Farmers of Southern Punjab face great challenges to effectively apply artificial intelligence (AI) for increasing crops' cultivation which result in declining overall crop yield followed by ultimate rising commodity prices in the country.
Agriculture is a major industry in Pakistan that contributes around 20% to the overall GDP and employs almost 50% of its’ population, directly or indirectly.
AI has the potential to help confront one of humanity’s biggest upcoming challenges of feeding extra 2 billion people by 2050 with only 4% of additional land will be available for cultivation - what has been projected by Pakistan and Gilf Economist (PAGE), a leading weekly financial magazine, in its report titled as Artificial Intelligence: The future of Agriculture.
These are the tools to assess operations coupled with tracking to reap its benefits such as decreasing crop waste, improving food security, minimizing the need for chemicals on crops and using resources sustainably.
According to the research of Sustainable Development Policy Institute (SDPI), presently there are three major applications of AI can be used in the country.
Firstly, the Agricultural Robots. It has the capacity to independently handle basic harvesting tasks at a faster pace with greater volume as compared to humans. This also includes efficient ways such as computer vision for farmers to protect crops from weeds without affecting the crop.
Secondly, a major use of technology can be in monitoring and data analysis. This will tackle many issues like degradation of soil quality that is a major contributor to food insecurity and have an overall negative impact on yield. Computers can use deep learning algorithms to store and process data captured by drones, GPS and field sensors installed in tractors etc. regarding potential defects and nutrient deficiencies in the soil and crop. This data can be used to monitor the health and readiness of crops and soil with regard to certain defects and diseases. It would enable farmers to take appropriate measures.
Lastly, predictive analytics also have a role to play in the form of precision farming. Machine Learning models can be used to track and predict external factors that impact the yield such as temperature changes, rainfall, wind speed and market shifts etc. For example, an existing app in Pakistan already provides weather notifications, whereas machine learning can take this a step further by customizing the predictions based on the needs of each client.
Challenges for Technology Adoption:
First of all, there is a lack of adequate crops data required for the application of such technology. In order to match up all different variables and their possibilities, the system requires a lot of different combinations of soil types, weather, seed variety, fertility, water availability, time period etc.
Secondly, it may be hard to sell such technology in areas where agricultural technology is not common. Farmers will most likely need help to adopt it.
Third, new technologies often seem confusing and unreasonably expensive because AgriTech providers fail to clearly explain why their solutions are useful and how exactly they should be implemented.
Fourth, one of the major hurdles is that Pakistan is in the very early stages of this revolution, thus, needs extensive testing and endorsement for these AI-based applications.
Fifth, it would also require farmers to be equipped with training to ensure they use the technologies effectively by understanding the data and insights provided by experts.
Sixth, a long-term plan combined with adequate funding is needed to test these technologies in the field.
The key to AI being accepted by everybody is the transparency of data and fairness in algorithms. These are the biggest issues that AI needs to tackle.
Commenting over challenges farmers are facing here, Assistant Professor of computer science from Muhammad Nawaz Sharif (MNS) Agricultural University having
expertise in AI subject, Dr Ayesha Hakim said, as an early player in this industry, Pakistan is struggling with data availability and documenting it accurately that is essential for training the algorithm and mining the data.
She said the university used to conduct training workshops for farmers hailing not only from Multan, remote districts of the Southern belt here.
She said it's a challenge to process and analyse web content, preparing audio, video files with creating easy text to educate farmers about the use of technology.
She admitted that there are financial constraints to the technology usage, thus urged upon the authority to 'subsidize the issue' locally.
Compared with New Zealand surviving with an agro-based economy, she anticipated that the local cultivators would need a decade to familiarise with the new invention.
As for as our neighbouring countries are concerned, Dr Ayesha said: India is far ahead of adopting AI as compared to other regional powers. She said Pakistan comes to the second position with regard to pursuing technology. Bangladesh and Nepal fall behind either to take or use the technology in domains of agriculture.
"It's in nowhere that 100% technology is used, but needs to have access and acceptance of the network by maximum people" said VC of MNS Agricultural University Professor Dr Asif Ali. It's not for everyone but the young farmers, particularly those who want to protect their fields.
Replying query about promotion, the VC said recently an MoU is signed between the university and Concave Agri, a European company, to hold advertisements about AI usage at a mass level.
Dr Ayeaha further said the university's management is trying to get two projects including Cacophony and Mosquitoes Surveillance from New Zealand to save our potential fields from flies attacks by calculating 'wind beat frequency' from affected areas.
According to Dr Asif, farmers are happy with pinning hope they would become familiar to the invention in two to three years. He said the university's experts has prepared an App. Just download it and put field data to have benefits, he said. Through App, farmers would be able to find out flies attacking their crops, he said and adding that they are facilitating farmers to gain the facility through every network they keep in their cell phones.
AI can also reach growers by sending them reminders and other important messages on cell phones. It can translate texts from different languages to the ones that they understand.
By collecting data on plant growth, AI can help produce crops that are less prone to disease and better adapted to weather conditions. In addition, farmers would be able to hold spray only at the targeted area where pests or flies are detected. It wouldn't only help them save money by avoiding spray to the whole of the field, protect human health from hazards of spray and the overall environment, she said.
Analyzing market demand, forecasting prices, and determining the optimal time for sowing and harvesting are key challenges farmers can solve with AI, she said.
AI can help judge insight into soil health, provide fertilizer's recommendations, monitor weather and track the readiness to produce crops, she added.
According to VC MNS Agricultural University, this is also a challenge for software companies. They should approach farmers gradually, giving them simpler technology first via agriculture trading platform.
Privacy and security threats like cyberattacks and data leaks may cause farmers serious problems. Unfortunately, many farms are vulnerable to these threats.
"There’s still a lot of work to be done by technology providers to help farmers implement it the right way," said the VC replying over privacy and security threats.
Government has the capacity to provide 'agriculture credit' to small farmers to enable them to invest in these inputs and techniques.
"Change is inevitable, whether you like it or not. Let's not give in to the fear of possibilities" said Dr Asif Ali.
He elaborated, a drone camera flies over a field to take images within minutes. The images will be used to forecast weather changing patterns, moisture, wind pace, soil condition, harvesting time and detection of crops' diseases in affected areas.
Modern spray machines with some particular sensors for detection and elimination of harmful weeds in the agriculture fields can be introduced by the dint of artificial technology in the country.
Associate Professor Dr Amjid Iqbal of computer science, University of Central Punjab, stressed the need for the joint working of agriculturists and IT experts for ensuring AI technology.
Dr Irfan Jafar, Head of Department of Software Engineering Superior University said artificial intelligence has been proved to be the basic reason for the progress of developed countries based on agro-based economies like New Zealand.
Dr Shoab Khan, Professor of Computer and Software Engineering in NUST said this regard that AI is believed to have USD 15.7 trillion impact on the global economy by the year 2030. According to him, AI and allied technologies are shaping the environment and life not only in agriculture, every field of life almost.
Currently, nobody is supposed to be heading the entire subject of Artificial Intelligence (AI) in the country. According to sources, the President of Pakistan has been assumed to be acting as head of AI subject and its overall operations carried out in Islamabad or elsewhere in the country.