Safety is of paramount importance to all rail operators and a core strategic tenet for each of the Class I Railroads in NA. Accidents and incidents are reported to the Federal Railroad Administration (FRA), and a review of recent accident data shows an overall improvement in accident rates of 10.7% since 2014. However, derailments remain the most common train accident with track issues a leading cause.
Size, complexity, and manual processes limit the opportunity for improvement
In North America, Class I railroads own and maintain their respective networks and infrastructure, including assets like track and bridges. The size of these networks varies from 6,000 route miles across 10 U.S. states and Mexico to an impressive 32,500 route miles across 28 US states. Each railroad has developed a maintenance of way program to inspect and maintain their broad networks to ensure safe and efficient operations. Typical inspection programs call for inspection of every foot of track up to twice a week. The complexity in planning and scheduling the people and equipment necessary to accomplish this as well as fully leveraging the volume of data is daunting, not to mention prone to error due to the manually intensive processes.
It stands to reason that more frequent inspections would likely lead to increased safety and fewer track-related derailments; however, due to the size and complexity of the networks and many sections of track being in remote areas, increasing the frequency of manned inspections comes at great cost and inconvenience.
An intelligent enterprise can help overcome these challenges
Organizations have access to a plethora of data, but it tends to be siloed with no single version of the truth. This makes analysis and extracting insights challenging if not impossible. In the digital world, organizations need to be able to turn data into insight into automated actions in real-time while seamlessly connecting their supply chain for execution. Figure 1 illustrates this premise and offers a digital transformation framework.
Figure 1: Digital transformation framework
Furthermore, Figure 2 illustrates how the elimination of manual processes and tedious tasks through automation can have a profound impact on an organization’s ability to be proactive and spend more value-added time and energy on what matters most: the customer.
Figure 2: Intelligent enterprises elevate employees to focus on higher-value tasks
Applying the digital transformation framework to rail
Let’s apply this framework to rail inspections and ultimately improving rail safety. It starts with access to rich sources of data, including sensor, maintenance, and even high-fidelity image data. Incorporating drone inspection technology into engineering maintenance programs for linear assets and structures like bridges enables an increase in the quality and quantity of inspection work that can be accomplished by having a direct correlation on safety.
For example, being able to capture high-resolution exterior shots of bridges as well as the ability to fly underneath and inside recessed areas, which are difficult and dangerous to manually inspect, ensures critical load-bearing areas are captured from every possible angle. LIDAR equipped drones can detect hairline cracks in rail and ties often undetectable through visual inspection. Additionally, beyond the visual line of sight and autonomous operations are now enabling more efficient and cost-effective drone applications. But capturing the data is only the beginning.
The critical next step is to turn inspection data, which has now been enriched with high-volume, high-fidelity image data into real-time insight. Leveraging image recognition capabilities powered by machine learning offers a robust and extremely efficient way to automatically analyze and identify potential maintenance issues. The system is trained to look for cracking, crazing, discoloration or other indicators that often precede a failure. Machine learning technology offers a very high success rate which will continue to improve over time as the volume of data increases and it continues to learn.
Leveraging a business rules framework, it is possible to then automatically initiate the required intervention. For example, images reveal cracking on a gusset plate connecting structural members of a bridge. The image is identified as being out of tolerance and in need of repair. A work order (WO) is automatically created inside the asset management application within the digital core along with a requisition for materials to complete the repair.
Finally, integration with suppliers and partners completes the end-to-end process. For parts not on hand in inventory, a PO is automatically created and sent to a preferred supplier via a fully integrated business network. And communication and scheduling of the physical work with full time or contingent maintenance of way gangs can be seamlessly accomplished through tight integration with a total workforce management solution.
Figure 3: Drone inspection seamlessly incorporated into an end-to-end process
Delivering the intelligent enterprise to improve rail safety
Adoption of drone inspection capabilities to supplement and/or replace traditional inspection methods as part of a seamlessly connected end-to-end process (Figure 3), has the potential to further move the needle on an already impressive rail safety record. Many railways are starting to embark on pieces of this journey today, but critical to long-term success will be leveraging an integrated platform to bring it all together in real-time.
Beyond safety, additional benefits could include improved productivity, increased network velocity, as well as higher asset utilization. This is only one example of how SAP is uniquely positioned to help companies turn data into insight into automated action and to deliver the intelligent enterprise.
For more on emerging technologies in business, see Creating The Intelligent Company.