Hub Group has recently announced enhanced end-to-end visibility, an innovation designed to deliver real-time, dynamic, shipment-level estimated-time-of-arrival. This is realized via a GPS-equipped container fleet, plus railroad data and a combination of artificial intelligence and machine learning technology.
To discover more, and to discuss control of goods more generally, Digital Journal spoke with Vava Dimond, EVP and CIO at Hub Group. The supply chain solutions provider is based in the Chicago area.
Digital Journal: What are the main challenges facing logistics and supply chain?
Vava Dimond: Even in today’s digital era, the biggest challenges logistics and supply chain companies face are often a lack of ability to collect and translate timely and accurate data. With the increasingly fast-paced nature of the supply chain, using data to inform insights to inform smart business decisions is critical.
Customers demand real-time updates received in other aspects of their daily lives, like weather updates and the delivery status of their most recent online order. To effectively evolve with increasing expectations for real-time insights, customer-centric logistics and supply chain players like Hub Group are working to fill those blanks with valuable intel using modern technology like data science and machine learning.
DJ: How is digital technology, in general, changing the process?
Dimond: In its simplest form, technology is making logistics and the supply chain more modern. Paper processes are now enabled digitally, voice technology can be used for data entry — essentially, technology is reducing the potential for error that can arise in traditionally manual supply chain processes. Our fleet drivers can now complete their jobs in an essentially mobile work environment.
DJ: What is the new technology developed by Hub Group?
Dimond:We have leveraged the very latest in AI and sophisticated machine learning models to analyze more than 10 million data points to deliver real-time ETAs. This allows for more accurate predictions for our customers. We’re also leveraging IoT devices on our fully GPS-enabled network, which allows us to use our extensive collection of data to leverage estimated time-to-ground (ETGs) predictions created by our machine learning models. These ETGs are now significantly more accurate, which feed our ETA engine to create accurate, timely delivery predictions, creating end-to-end visibility for our customers and those they’re doing business within the supply chain.
Updates within the system occur dynamically based on transit events and truck speeds, as well as traffic and weather conditions. These events are continually processed, analyzed and updated at the rate of 1,000 predictions per hour. We also leverage cloud technology, which allows us to auto-scale the model on demand.
The technology covers 100% of Hub Group’s intermodal network, including both local (shipments handled by one railroad) and interline (shipments handled across multiple rail ramps) lanes. Our testing indicates that our models are extremely predictive and accurate. This enables us to better serve our customers and deliver the full transparency and visibility they demand.
DJ: What were the main challenges when developing the technology?
Dimond:Collating both timely and accurate data to program our AI models and output useful insights for our customers to easily access was the number one challenge. The key is understanding which information was most meaningful and why when sifting through the mass amounts of data in developing this technology.
Our end goal was to provide our customers with the best end-to-end solution possible; with that, we were careful and meticulous in what we looked for in that data and fed into our systems. Collecting the data to be most useful in predicting ETAs was no easy task. By iterating on this approach with great attention to detail, we were able to develop a solution that provides our customers with the most accurate and informative look into the supply chain they could want. We couldn’t be happier with the end result of the product.
DJ: How accurate is the technology?
Dimond:Our models analyze more than 10 million data points – among them include all of the data coming from our GPS-enabled fleet, information from the rails, truck speed, traffic and weather conditions. All of these inputs are continually processed and updated in the system. It’s that dynamic!
We believe the accuracy is at a point that it can drive better decision making in the supply chain. With AI and machine learning implemented into the solution, this data gets stronger over time, which allows us to continually meet the evolving needs of our customers and stakeholders.
DJ: How will operators use the technology?
Dimond:Knowledge is power, and when you have a complicated supply chain, knowing the estimated arrival time — whether that be early, late or right on time — can help sophisticated shippers keep their supply chain moving efficiently and running related business functions optimally. The benefit of Hub Group’s technology for operators is being able to have visibility throughout the lifecycle of all shipments and having the ability to make adjustments based on that real-time information.
For operators, it’s often as simple as knowing what to expect in the supply chain. Providing an ETA allows the operator at a delivery point to know just that. Armed with this information, operators can make the best decisions coupled with the many other data points coming in from the supply chain to inform their day-to-day.
DJ: What types of data will operators be able to draw upon?
Dimond:With Hub Group’s new end-to-end visibility tool, operators will now be able to draw upon the more than 10 million dynamic data points that we gather to be used in the calculation of ETAs for our customers, including transit event information, truck speed data and traffic/weather condition data. In addition to the ETA, the operator now has access to all of the “events” occurring with a shipment, which at Hub Group, we refer to as “Shipment Details.” This includes all events when a driver has been assigned a delivery that goes into the calculation, which ultimately provides operators with the most holistic view of each shipment.
