Troika Solutions was recently invited to participate in the United Kingdom’s Ministry of Defence (MoD) and Security Accelerator (DASA) Logistics Hackathon, conducted 29-30 November 2018 in London, England. The Hackathon focused on accelerating logistics decision support through exploitation of ML capabilities. Troika Solutions was asked to participate in this event based on demonstrated data sharing and data fusion expertise coupled with Troika Solution’s expansive work in Logistics 4.0, which is rapidly becoming recognized as the digital revolution of business. Logistics 4.0 seeks to expand the supply chain from a linear depiction to a network of processes and technologies that improve planning and execution. Logistics 4.0 includes machine-to-machine communication, autonomous systems, digital manufacturing, and iterative decision-making. This supports highly improved forecasting, resourcing, and distribution that improves responsiveness, increased velocity, and sources of support all while reducing footprint. Logistics 4.0 is sometimes referred to as smart logistics. Smart logistics requires technologies to identify, sense, locate, process, decide, and act. Machine Learning (ML), a key element of Logistics 4.0, is a method of data analysis based on the idea that systems can learn from data, identify patterns, and make decisions with minimal human intervention.
The intent of the Hackathon was to demonstrate the ability to analyze and share structured and unstructured multi-source data while maintaining its classification and permission based access rules at machine speed. Data sets from the C-130J Hercules aircraft were provided to enable the development and testing of potential data sources.
The Hackathon rules required over 70 attendees to form into 11 teams within the first 30 minutes of the Hackathon. The teams then had 18 hours over two days to individually collect, collate, cleanse, and control the provided data in accordance with established business rules. A working proof of concept was developed and demonstrated, and a white paper describing the team’s approach and architecture was delivered. Troika’s Director of Integration Services, Mr. Kevin Clyde, and Director of Portfolio Management Services, Mr. Jerry McGovern, both participated in the DASA event. Troika teamed with Diem Analytics, PA Consulting, Massive Analytics, and the University of Central London. The team used machine learning to clean and enhance the data with metadata so all data feeds could be collected, collated, cleansed, and controlled. ML was used to predict major trends in readiness for platforms with high usage. The approach included the use of open source technologies to create a horizontally scalable platform and iterative processes that prepare the data per the Hackathon rules.
The iterative architecture model Troika proposed was designed to initially establish permission-based access rules at machine speed. The second iteration uses the same mechanics to identify trends and patterns related to operational considerations. This enables algorithms to begin providing insights for accelerating logistics decision support. Subsequent iterations continue in order to allow machine learning to deliver outcomes such as predictive maintenance, inventory optimization, and product life extension. This approach is depicted in Figure 1.
Troika’s Hackathon experience directly integrates previous work targeted at Reliability Centered Maintenance (RCM). RCM specifically provides a recognized and qualitative means to transition a rigid preventative maintenance program to a more agile and accurate predictive maintenance program.
This transition is fundamental to supporting forces in austere operating environments, optimizing enterprise resources, and managing costs, all the while improving readiness and ensuring operational success.
Adding ML to an RCM approach not only accelerates logistics decision-making, it greatly increases human capacity to calculate and address the spectrum of variables that impact operational success. The precise data required for the algorithms to support machine learning focused on RCM, also supports and complements the algorithms to optimize repair parts inventory and product life extension decisions.
At Troika Solutions, we pride ourselves on being able to find creative ways to integrate the relationship between operational needs and innovative technology, especially for customers that rely heavily on complex data management for decision-making. We make data more useful, visible, accessible, and actionable – via a portfolio management (PfM) approach to solutions and capabilities. We are a trusted partner for decision support solutions that leverage data fusion, visualization, and analytics to create “Built for Purpose” tools that provide increased data interoperability and velocity for critical decision making. At the DASA Hackathon, we demonstrated our ability to add a Logistics 4.0 technology such as machine learning to the “built for purpose” tool box.
Troika Solutions is a Small, Disabled, Veteran-owned Business. For more information, contact Kenneth Lasure by email at email@example.com, or by phone at (571)-375-7142. Find Troika Solutions online at www.troikasol.com.
Edited by Sarah Plaut