How ML and AI can give FedEx early warning of problems
FedEx has applied advanced machine learning across its operations by adopting tools in Microsoft Azure such as Azure Databricks, Azure Machine Learning, and Azure Data Factory, as well as external solutions that easily integrate with Azure such as GitHub. rice field. Anthony Norris, senior IT vice president for global platforms and customer solutions at FedEx Services, said machine learning could take data intelligence to the next level, acting like a kind of canary in a coal mine, helping organizations It explains that it alerts leaders to challenges and problems. Separate delivery available.
“Machine learning can help predict if a package has gone off track and intervention is needed,” Norris says. “This is a great value proposition for our customers.”
At a higher level, the system helps FedEx uncover broader issues. For example, during the 2022 holidays, issues related to deliveries requiring customer signatures accounted for 25% of calls to FedEx call centers, but only 2% of deliveries required such signatures, Gibson said. said. “That experience showed us that something was wrong, and we have a team that can take that insight and improve our operations,” he says.
expedition: What impact will AI have on cybersecurity in the next few years?
In fact, FedEx was inspired to create internal teams focused on different issues based on data collected from machine learning. One, for example, focuses on solving the challenges of time-sensitive pharmaceutical delivery. “AI helped develop so-called package his fingerprints,” says Gibson. “Based on that data, we created a monitoring and intervention team and a dedicated dashboard to track high-priority packages.”
Before the advent of machine learning, Norris said FedEx’s data intelligence collection methods were “rule-driven.” “However, when the network is complex, it is difficult to layer the rules to cover all scenarios.”
Machine learning models can adapt to dynamic situations in real time and apply what they already know to the problem at hand, allowing companies to build more intelligence into their models. “For example, the model alerts us and tells us when and what we need to communicate with our customers. Much faster and more accurate,” he says.
Norris and Gibson agree that increasing FedEx’s credibility and visibility will benefit customers.
“The benefits of machine learning are multifaceted,” says Norris. “We can identify root causes of problems and opportunities for automation. AI and machine learning can really help us do that.”
