AgriScan (Team 21)

Plant Scanning Page
Text Messages
Map Page
Result Page
Home Page

Inspiration

The growing global population, which recently surpassed eight billion people, presents a significant challenge to our food production systems. However, food production levels are not keeping pace with this growing population. Climate change and global warming are transforming once-arable farmland into barren wastelands, where crops can no longer grow. These same temperature changes are also favorable for microorganisms like bacteria and viruses, which can thrive in such environments, causing considerable damage to crops.

The ongoing challenge of feeding an ever-growing population in the face of dwindling resources and increased disease outbreaks is a problem that requires immediate attention. Although combating global warming is a long-term solution, it won't be enough to meet the immediate needs of our food production systems. Therefore, we need short-term solutions to increase food output.

Sadly, every year, plant diseases and pests account for the destruction of 40% of global crop production, a staggering figure that is unsustainable in the long term. Without effective measures to preemptively detect and eradicate these pests, the destruction of crops will only increase, leading to food shortages, famines, and skyrocketing prices.

The good news is that there are technologies and solutions available that can help farmers and organizations to detect and monitor crop diseases, making it possible to take proactive measures to prevent their spread. The use of advanced geospatial data, combined with innovative disease detection technology, can provide farmers with a comprehensive view of the spread of crop diseases globally, allowing them to make informed decisions about their crops.

By using these tools, farmers can stay ahead of crop diseases, take preventive measures, protect their yields, and, ultimately, contribute to feeding the world's growing population. At the same time, agricultural research organizations can use this data to identify patterns of disease outbreaks and find new ways to combat these pests effectively.

In today's world, everyone has phones, and a network to connect with others. The widespread adoption of mobile phones has created new opportunities for global collaboration and knowledge sharing, particularly in the agricultural sector. By leveraging the global farmer community, we can harness the collective knowledge and expertise of farmers from all over the world to address the challenges facing the agricultural industry.

Source: https://www.fao.org/news/story/en/item/1402920/icode/

What it does

AgriScan is an online platform for farmers (or anyone near crops) to find and identify diseases in their crops. It utilizes complicated neural networks, with datasets of thousands upon thousands of images. The application is designed to be as simple as possible for people all over the world. The platform does not have a user authentication system and it contacts users through their cell phones to make it as easy as possible to use. Once the user enters their phone number, location, and crop image, the application identifies whether the crop is healthy or diseased (and if so, what disease). The user's data is also logged as geospatial data and put on a global map that integrates a hotspot algorithm so farmers can easily identify whether an outbreak is happening in their area and work towards mitigating it.

How we built it

We built the platform using Flask and Python to allow for rapid prototyping and development, as well as simplicity and ease of use to onboard new individuals to the team in the future. A PyTorch model with a public plant disease dataset was used to ascertain the disease a plant has. The website was posted on PythonAnywhere, a cloud-based Python web app deployment service.

Challenges we ran into

The hardest challenges we had in this hackathon were the time limit as well as our unfamiliarity with the technologies. The 24-hour time limit was stressful and limited how far we wanted to go past the MVP of the product. Our group also did not have any experience working with and storing/managing geospatial data.

Accomplishments that we're proud of

We all learned how to use geospatial data with maps and managed to work together despite not having formed our team before the hackathon. We also managed to get an MVP developed rapidly and had time to spend on adding features.

What we learned

Not only did we learn technical skills, we learned about development practices (GitHub, lining, merging), and how to put together a team on the fly and still work well together.

What's next for AgriScan

The Avian Flu is one of the most devastating animal-borne diseases in bird history. We hope to expand AgriScan to the IoT sector as well, as a network of cameras around chicken farms could notify farm owners if any birds flying abnormally (meaning they have the flu) are present in the area. AgriScan could also evolve to the extremely advanced early detection of diseases/pests through soil NPK (nitrogen, phosphorus, and potassium) sensors that can identify if certain crops are susceptible to being destroyed.