Maersk participates in a food waste project with Wageningen University and several of our customers. The end goal is to create a prediction model for food freshness and remaining shelf-life.
Each year, 13% of the global fresh produce is lost in the cold supply chain, between harvest and retailer, which equivalents millions of Euros in revenue. With an accurate, real-time prediction of remaining shelf-life of ap articular shipment of fresh produce, it would enable all parties involved in the supply chain to take earlier and better-informed decisions, ultimately reducing food waste and avoiding lost revenue.
This is exactly the challenge that a new project is aiming to solve. Maersk participates in the project together with Wageningen University, a public research university in The Netherlands and among the top 150 universities in the world. The end goal is to create a prediction model for food freshness and remaining shelf-life determined by multiple quality parameters.
The project was initiated earlier this year and will be driven by the Wageningen Food & Biogen Research, on the research institutes of Wageningen University & Research, in a consortium with 13 private partners from different parts of the fresh food supply chain.
Creating a Digital Twin
According to Ken West, Reefer Digital Development Manager, predicting cargo outturn has for a long time been an area of interest for the Reefer Solutions department in Maersk.
“Today we have a great amount of data for all our reefer shipments from the Remote Container Management system, and we have tried to use this data to predict cargo loss and risk of claims. But in order to give a proper prediction, we simply need to know more about what happens to the cargo before it’s stuffed into a Maersk reefer”, he explains.
That’s what we are hoping to get from this project, now that we have players from across the entire cold chain involved.
Today we have a great amount of data for all our reefer shipments from the Remote Container Management system, and we have tried to use this data to predict cargo loss and risk of claims. But in order to give a proper prediction, we simply need to know more about what happens to the cargo before it’s stuffed into a Maersk reefer.
AgroFair and Westfalia, two current customers of Maersk, are also part of the project consortium and the partners have started sharing data sets with each other, ranging from details about weather and soil at the harvest farm to qualitative evaluations of product samples and specific temperature readings.
It is the intention to combine all relevant data and create a so-called Digital Twin. A digital twin is basically a digital representation of a physical object with underlying models that can simulate the real-life behavior of, int his case, fresh produce. If the Digital Twin is connected to real-time information, it can even provide real-time predictions, and the predictions can be constantly improved based on a feedback loop.
A future mission for Captain Peter
The project could in future also benefit users of Captain Peter, our digital assistant for reefer customers.
If we are able to come up with a prediction model that works for different reefer commodities, it would make sense to integrate it with Captain Peter – this would fit perfectly with the Captain Peter vision of being the digital assistant who is handholding our customers through the reefer journey and always giving the best advise based on what is known at that point in time.