ComplEat

Inspiration

The Covid-19 pandemic has created massive disruptions in the food system. In some cases, we have observed and experienced severe food shortages. In others, food cannot reach end consumers and is ultimately wasted. There are many factors that contributed to waste, including: market factors, products not meeting aesthetic standards for a buyer; or damage from weather and pests. Yet one factor significantly increased waste during the Covid-19: miscommunication between producers and retailers.

Particularly, one-third of the food produced globally for human consumption, corresponding to about 1.3 billion tonnes, is lost or wasted each year along the food supply chain. But it’s not just the food that is wasted, but also the number of labor days, the gallons of water, and the thousands liters of fuel, just for products that sit in landfills and produce C02 emissions.

What it does

Our mission is to help prevent food loss and waste along the supply chain. In particular, we ask ourselves: How might we help food traders optimize their storage and supply chains so that food processing companies and retailers find the raw materials they need in a safe and accessible way?

How we built it

We created a tool that combines AI for shelf life prediction, dashboard for tracking and a digital marketplace to connect traders with potential customers based on their demands. In particular, The imaging software determines what the item is, whether it has any damages and how many more days it has before it rots. This data is then visualized on a our website dashboard for traders to track what is in the warehouse and sell the produce to the right buyers. The freshest produce for example will be sold to supermarkets, but less fresh fruit can still find uses elsewhere, where it can be processed or used immediately.

Challenges we ran into

We ran into multiple challenges. For the designers, we have to figure out the user flows for both the traders and the buyers' perspectives. We were not able to get primary research such as interviews or surveys with our target users, instead, we gained insights from multiple secondary research sources including reports, news articles, research papers, etc. For the developers, we have to learn by ourselves a lot of new frameworks related to AI/ML algorithms in just one day, particularly image recognition and prediction.

Accomplishments that we're proud of

At the end of the journey, we're proud that we have successfully created a running website with all the features we initially came up with: including the dashboard for tracking and a digital market place that automatically matches traders with buyers based on their demands, fueled with data we got from our AI/ML algorithms.

What we learned

We learned to communicate more effectively to each other, and reached out to mentors as well as the Dubhacks community when we ran into any difficulties.

What's next for ComplEat

At the moment, our AI/ML algorithms only works best for certain types of fruits and vegetables, thus, we're hoping to expand our tools for a diverse range of fresh produce. We're hoping to conduct multiple pilots where we can help traders find new homes for foods that are close to their final use-by date.