Air traffic is increasing rapidly. IATA has mandated resolution 753 that requires bags to be scanned at 4 points: check-in, aircraft loading, transfers between aircraft and on arrival. However, seen from a security perspective and the core purpose of BHS, it is also necessary to keep track of the bag through the whole flow in the BHS. By tracking bags through the whole Baggage Handling System, performing continuous analysis and prediction, operators of airports can predict heavy baggage flow, if the system requires maintenance or is appropriering breakdown states. To accomplish this, a number of classification and prediction algorithms needs to be developed and tuned, AI methods needs to be investigated and neural networks need to be developed and trained. To improve situational awareness a dashboard that provides operators with a real-time status and salient predictions, need to be developed and fine tuned to reduce information overload. The big data platform provides the basis for preemptive and predictive capabilities and it need to transform and merge data into various applied statistical time series analysis, neural networks and supervised/ reinforcement learning. Among the hardest of the challenges, is how to apply video-data for inventory and flow control and how to properly distribute and optimize Agent responsibilities in the Multi Agent System. Assuming a successful project, airport operators will be able to leverage existing data to decrease congestion, reduce cost of maintenance and breakdowns, energy use, personell costs and to more efficiently employ existing BHS eliminating needs for upgrades. Airlines will be able to better coordinate the flow of baggage to reduce delays on tight of delayed transfers in accordance to the principles of priority and fairness. They can also use the data platform to provide their customers with fee-based tracking of baggage.
Project leader: Espen Remman
Institution: SENSEC SOLUTIONS AS