Silicon Valley is Helping the Logistics Industry Grow Smarter, Instead of Bigger
Q: How can shippers and service poviders cut through the clutter of tech talk around AI and machine learning?
A: The most important thing to remember is that machine learning is not a generic box you can buy from just anyone, plug in and have work. The shipping industry is so complex and nuanced that any AI or machine learning solution must be custom built for logistics for it to deliver scalable value. Also remember that machine learning and AI solutions are not all created equal.
Q: Why is there so much interest from Silicon Valley in the logistics industry—and vice versa?
A: In short: because the two worlds can really help each other—and there's such massive value to be gained through partnership. We see incredibly talented and experienced supply chain professionals who are unfairly burdened and not equipped with the best tools.
The machine learning and AI used every day in Silicon Valley can be of great help, especially at a time where economic pressures are requiring innovation beyond the four walls of the supply chain.
The logistics industry has spent years throwing capacity and scale at its problems. Shippers and service providers have added bigger ships and ports, capacity and buffer stock. But bigger no longer means better. Enormous pressures are weighing on the industry, forcing it to be more efficient in operations and assets.
A tighter handle on the movement of goods is needed and the only way to get smarter and more efficient is through data intelligence (versus scale). A select few companies are customizing Silicon Valley data science for supply chain. They are delivering high value to supply chain operators—and therefore gaining rapid traction.
Q: Where is there value being captured today?
A: One area of low-hanging fruit is the deployment of machine learning to cleanse and sort data to prepare it for greater simulation and analysis. The state of data today is an inhibitor to innovation. There is immediate value in using machine learning to solve the data problem.
The sexier use cases for AI and machine learning include predictive logistics and things like PTAs (predictive time of arrival) to know four or eight weeks out what will happen with regards to shipment movement or trading partner decisions. The shift from guesswork and estimations to data-driven predictions will be the most impactful result of AI.