Analyst Insight: In the wake of recent supply chain turmoil, resiliency has become a key element in the planning, development, and execution of logistics operations. For shippers who are dependent on single-mode logistics, the time has come to take a close look at how multi-modal services can be added to the options that will keep things moving when unexpected glitches develop.  

Not only is it critical to develop these options for supply chain reliability, but also in order to achieve safer, sustainable, and more efficient business models. 

Rail shipping can provide these attributes, recognizing that it comes with a set of variables that must be managed to drive the success or failure of shipment delivery. 

The best way for shippers to deliver safely, sustainably, efficiently — and with resiliency — is to use data that supports accurate and near real-time shipment visibility. You are best prepared to fix problems when you know where and when disruptions might occur. 

Ultimately, the goal is to provide your customers with a solid estimated time of arrival (ETA). Historically, rail ETAs relied on moving averages or other simple models, without accounting for other possible operating conditions that could affect transit time —including service days, differences in train types, and consistent delay trends — leaving customers with a less-than-optimal ETA.  

Also, shippers lacked an integrated single-source of rail shipment status data, instead having to piece together disparate data from multiple railroads, third parties, ports, and first/last mile carriers. 

Now, advances in rail industry technology have created integrated tools that put quick-action solutions at your fingertips. These tools are rapidly evolving with the use of Artificial Intelligence (AI) and machine learning to create dynamic predictions based on thousands of origin-destination pairs. These models analyze and process information much faster and deliver it to the shipper in near real-time. 

Utilizing sequence modeling, the system quickly learns over the course of trips how to identify the most important sequence elements and other real-world factors that will impact a shipment’s arrival time. This provides the ability to predict and update expected ETAs in near real time as shipments move over the rail network. 

This rich data stream enables rapid adjustments to regular operations and supports development of out-of-the-box ideas that keep distribution centers properly stocked, and plant operations efficiently producing goods. 

Sequence modeling has greatly improved during the past three years. Today, the rail industry can provide:

• Near real-time visibility.
• More reliable shipment ETAs.
• Enhanced historical analyses.
• Alerting and exception management.
• User-configurable dashboards with customized insight capabilities.
• Map visualization of shipments.
• Analytics and insights powered by AI and machine learning. 

The work to improve rail visibility has yielded significant performance improvements, with development continuing. Data scientists are working on location sensors, geo-fencing, and network health and performance maps to create visual data. Functioning across the various applications, users will be able to access this visual data to create metrics, and interact with predictive and prescriptive analytics. 

Outlook: Recent improvements in technology have convinced many shippers to add rail into their transportation mix as a greener and safer option. Shippers can develop strong, resilient supply chains that incorporate more rail service, which is safer, more sustainable, and more efficient than single-mode highway service.  

Greater use of multimodal solutions will allow shippers to “have it all:” long-haul efficiency, flexible last-mile delivery, and more shipping options in support of supply chain resiliency.