This post has been republished via RSS; it originally appeared at: New blog articles in Microsoft Tech Community.
The Digital Ag Hackathon hosted at Cornell bought together 250 students from Cornell, the University of California Davis, the University of Sao Paulo from Brazil and Wageningen University from the Netherlands. These students came with a single mission - to spend the weekend trying to solve some of the problems faced by farmers as they try to feed a growing and changing world.
This was one of the most diverse hackathons I've been to, with a wide range of students across all disciplines from veterinary medicine, to plant science, to information systems. Each bought a unique perspective and background not often seen in a pure CS focused event. Due to this the solutions were less technical and did not all have working prototypes, but included a lot more focus on mock ups, the business plan and how these would be delivered. This meant that the problem being solved were much wider, with a real plan for how to convert these solutions into a finished business.
A group of us from Microsoft were on site to help with technical expertise, provide our thoughts and feedback on the event, and obviously bring copious swag including stickers to adorn the students laptops.
Azure Farmbeats is a purpose built solution accelerator built to allow farmers to get actionable insights from data gathered from sensors or drones on their farms, and external data such as satellite and weather data. Data inside Farmbeats can be sent to external services to allow more value to be obtained, for example sending it to Machine Learning models to make predictions on plant growth.
We came with some Farmbeats experts, including an actual farmer and member of the Microsoft Farmbeats team Andrew Nelson who uses Farmbeats to manage his farm. Students were introduced to the technology via workshops, and hardware kits that they could use to generate data.
You can read more on Farmbeats on the Microsoft Azure blog. Also check out this video starring Andrew talking about how he uses Farmbeats on his own farm.
Azure IoT and AI
To help the students learn about how the cloud could power the hacks, they were introduced to the services available from Azure via a couple of on-site workshops.
One workshop covered out IoT offerings, showing how they could gather data from sensors connected to an IoT device, and send this to the cloud for processing, including using this data combined with weather data to determine if a plant needed watering. This workshop is available as a hands-on lab that you can do at home with some cheap IoT hardware at aka.ms/AgroHack.
The other workshop covered our AI services, showing the power of Azure ML Studio to quickly create Machine Learning models using drag-and-drop based tools and no code. This workshop showed the same content we've taken on tour globally at Microsoft Ignite | The Tour and is available at aka.ms/AIML30.
The winning teams
Like every hackathon I've been to, all the students were able to put together and present some pretty amazing work in only 24 hours. In the end though, these events are a competition and there can only be a few winners. This hack had prizes for four different categories, as well as one overall winner.
Overall winner: AgPal
Team AgPal put together a solution for farmers in India to be able to get insights into the best markets to sell their produce to ensure that they are not only able to sell to that market, but to also ensure they get the best price. This app is designed for NGOs to track prices and stock, and make recommendations to the farmers via low-tech methods such as text messages.
Best Data & Analytics: MicroSoil
Team Microsoil put together a soil monitoring system using test strips and a mobile app to quickly get an overview of pathogens in soil, instead of waiting 6-8 weeks for the results of samples sent to a lab.
Most Novel: Bagasse
Team Bagasse made a prebiotic drink using the waste products from Tequila production. They actually made the drink during the event and it was available to taste, tasting slightly like tea.
Most Market Ready: ConsumABLE
Team ConsumABLE built a mobile app for folks with food allergies. This app connected to public APIs from stores to allow it to check for allergens using the products barcode. It also made suggestions based on reactions other app users had to foods.
Societal Challenge: Hummingbird
Team Hummingbird designed robot-based detection of plant diseases using cameras and AI.
If you want to learn more about Azure IoT and AI services, the best place to start is Microsoft Learn - our free on-line, self guided hands on learning experience.