– Develop systems to unlock sensor/data impact across value chain and enable disruptive Just-in-Time (JiT) feed supply chain. Real-time
simulations and heuristic models responding to new data (live silo data, feed plans, raw-material-supply data, external data/weather).
– Design, modeling and Implementation of Digital-Twin-System (DTS) of feed supply chain with stakeholder specific applications delivering data/model-driven insight.
Combining symbiotic simulation with optimization models considering uncertainty aspects during the planning horizon and extendthe simheuristics algorithms to solve the optimization problem.
Control of feed stocks in livestock farms is managed by the farmers in a manual and inefficient way. Most of them visit the silos one or two times per week and hit them with a mace to acoustically guess the stock levels. When the silos are empty the farmers send a refilling order to the feed supplier, who must manufacture and deliver the feed, often in less than 24 hours. This method forces feed suppliers to work on-demand, unable to organize and optimize their production or logistics in an optimal way. Furthermore, the lack of an accurate method to measure the stocks in the farms leads to necessary additional costs due to urgent orders caused by run outs and orders with wrong quantities. These inefficiencies cost the feed and livestock industries more that £2 billion each year at a global scale.
The thesis aims to develop a Digital Twin System (DTS) of the livestock feed ecosystem and a Smart Feed Logistic Platform (SFLP) a system that uses Internet of Things (e.g. sensors) and data analytics to enable real-time decision making in livestock feed supply chains, linking manufacture and supply with farms and nutrition advisors.
In SFLP, key data from farms (silo inventories and feeding plans) will be captured to feed our DTS platform and will be used to model a clustered feed demand function. Suppliers will establish Just-in-Time (JiT) agreements with their customers and will assume the responsibility to manage the inventories of the farms’ silos. The nutritional advisors will also take part, accessing farm data and helping the farmers to prepare the feeding plans and monitoring the daily consumption to introduce the required adjustments. The real-world data will come from industry first, patented, INSYLO 3D-Camera sensors, installed on farms, that monitor feed volume/temperature /humidity in real-time from silos, storing data in the cloud.
A symbiotic simulation (or digital twin) will be created to represent the real-world system virtual and simpler.
The symbiotic simulation will be combined with an optimization models while considering the uncertainty during the planning horizon, extending the simheuristics algorithms to solve the optimization problem. Outputs should include cost, risk and carbon foot print.