The airfreight industry is undergoing a transformation, with technology and artificial intelligence reshaping air cargo.
One of the key benefits of AI in air cargo operations is its ability to manage the complexities of supply chain issues swiftly and efficiently. AI techniques allow for the creation of a digital representation of the supply chain, encompassing production, procurement, warehousing, logistics, and more.
Using an AI system, a digital representation of the supply chain can be created, representing, for example, production and procurement, geographically distributed warehouses forwarding stock to where required, and logistics with time and cost implications. The AI then reasons how the supply chain performs given the demands from air cargo operations and delays or costs at any point in the chain.
“An AI approach brings faster and more accurate answers,” Simon Miles, Head of AI at Aerogility, explained.
READ: Transforming operations with technology
Forecasting and focus
Aerogility’s model-based AI technology helps with forecasting maintenance requirements for airfreight. This approach allows the system to simulate future scenarios, including maintenance schedules and availability. Unlike traditional machine learning (ML) methods, model-based AI can simulate the cumulative effects of multiple events.
Aerogility supports customers on a day-to-day basis in designing schedules which limit costs of maintenance, ensure minimum availability throughout a multi-year period and determine when best to upgrade aircraft. It uses model-based AI in which a customer’s knowledge of their operations and goals are used to play out simulations of future years.
“Compared to ML approaches, it has two distinct advantages,” Miles stated. “First, while a ML system can detect and extrapolate patterns in past data, model-based AI can present the consequences of many smaller events leading to larger effects, for example, the cumulative effect of aircraft being queued over time waiting for maintenance.”
“Second, model-based AI is inherently more transparent than ML. If your forecast presents something distinct or unexpected, such as a dip in availability, Aerogility allows you to dig into the reasoning behind why you are seeing this result and what you might do to change it.”
READ: Innovative digital solutions for airfreight operations
Enhancing decision-making
Aerogility’s scheduling algorithms are designed to maximise availability and minimise cost and ground time.
The simulations created in Aerogility can be used to conduct what-if analysis of various planning parameters such as cost, maintenance, repair, and operations availability or ground time variation to select the optimal schedule.
For example, planners can see if there is a period where a lot of annual leave is being taken that may impact operations or anticipating periods of higher demand.
“The key for being prepared for unpredictable events is understanding the robustness of your operations in advance, adjusting strategies and resource deployment to be ready for anything plausible,” Miles said.
“Aerogility first helps by modelling the complexity to give predictions accounting for that complexity. Moreover, it allows a potential scenario to be run, altered and run again in minutes to see the effects on long-term key performance indicators. This means that a wide range of possible futures, whether anticipated or not, can be tested in a timescale which allows you to prepare.”