Over 17 million hectares of land in Indonesia are used for coal mining, contaminating adjacent farmland with heavy metals and thereby reducing crop yield by up to 50%. Land rehabilitation is possible by executing phytoremediation, a process which uses certain crops to extract heavy metals from the soil. However, many farmers do not understand which plants are best for their land. A team of students majoring in a diverse mix of subjects such as Banking, Business Analytics, Finance, Marketing, and Food Science at the University of Sydney came together to form team Phyto. Their solution uses geospatial and weather analytics to design a tailored phytoremediation plan for each farmer. Based on a farmer’s location, the app uses IBM Environmental Intelligence Suite to access satellite images, which are then analyzed to determine contamination levels, soil conditions, moisture levels, and nearby vegetation. A Watson machine learning model then uses IBM Weather Data APIs to access these soil conditions and select the most suitable phytoremediation crop for the farmer to plant. For the farmers, analysis is all done remotely, reducing the need for them to do manual soil analysis on site. Additionally, once the plants are in place, the app includes growing guidelines, weather alerts, and land rehabilitation resources to help ensure the success of the new crop. Users can also use the “connect” feature to engage with community members and experts. Lastly, the “sell” feature provides access to an online marketplace which connects farmers to small local businesses who are sourcing raw materials. With little to no coding experience, the Phyto team was able to learn Watson Studio quickly to build their solution. They hope their solution can also eventually be used for larger scale land rehabilitation projects.

Approximately 1 out of every 3 people in the world does not have access to adequate food. Food waste also contributes to greenhouse gas emissions. The Synergy Squad team from Persistent is hoping their solution can address food insecurity and reduce food waste by making sure less food purchased by consumers goes to waste. Their “Offshelf” solution enables users to track the food in their homes, receive expiry notifications, and share surplus item within the community to reduce the chances of good food going bad. AI recognizes fresh items from a photo or voice description and provides an estimated expiration date, or packaged items can be scanned to input the date. Push notifications would let users know when groceries are about to expire, reducing the likelihood of wasted food. When users cannot make use of their groceries, the solution also includes a platform to facilitate the sharing of surplus food within the local community. Watson Speech to Text is used to help users fill out forms in the application through voice input. The team hopes to add to Offshelf, with additional image recognition support to read expiration dates and other details from food labels faster.