Program Manager responsible for ITIL processes for a major maritime shipping company with multi-ocean and country operations. Design, develop, and coordinate the integration of people, processes, and technology to improve the effectiveness of IT incident, request, and change management across multiple lines of business.

Designing Machine Learning Models and implementations to gain benefits of image analysis while minimizing the complexity of overall systems.

Responsible for developing and deploying integrated systems of robotic platforms to increase CSX capabilities and improve the corporate operating ratio. CSX is poised to take advantage of state of the art in robotics systems to meet the challenging needs of all-weather, outdoor operations on complex terrain. Starting with the mobile robots for yard operations, CSX plans to expand robotic systems to meet diverse needs in rail yards and other industry sites. Through a variety of robot platforms and payloads, CSX sees the opportunity to increase capabilities and reduce dwell times in rail times.

Program manager for an initiative to migrate all of CSX’s mainframe operations to a new internal cloud. Key to the initiative a process that migrates all active legacy code and databases to new services based on Java and PostgreSQL. The scale of thousands of new web applications requires a new infrastructure that leveraging both open sources and commercial software to deploy an integrated cloud environment that meshes a complex set of existing mid-tier applications to provide continuous service to the business. The initiative’s scope includes the dramatic expansion of cloud services to support thousands of new web applications. The rapid conversion to cloud operations at scale requires extensive automation with available tools including Bamboo and Jenkins. Completing the testing and implementation automation requires updating quality assurance processes to support automated testing using Jerkin/Cucumber. Critical to successful and timely build processes is the inclusion of test data management (TDM) to validate new builds properly. Concurrent quality assurance testing for hundreds of inter-related applications requires TDM to provide the ability to reset test data between the necessary revisions. Implementation and sustainment of extensive cloud operations require changes in organizational responsibilities to integrate development and operations into DevOps model across the organization. The shift in responsibilities requires active planning for the organizational changes required for integrating operational staff with mid-tier and mainframe development teams. Current mainframe developers have critical knowledge in mainframe systems and more importantly in the business needs behind the legacy code. Retaining experienced staff requires extensive planning for training all personnel, especially legacy experts, in new tools, techniques, and systems.

Managed the the development and implementation of a complex image analysis system to automate the daily inspection of thousands of railcars. The pilot site used a combination of 18 line and area scan cameras to collect imagery of cars transiting portal. The system design supported a combination of computer vision and deep learning in a library of hundreds of algorithms that will run on tens of portals. The project worked with business experts to establish the critical check conditions based on the multiple types of car, load conditions, HazMat constraints, and operations checks. The initiative design used a combination of vendors targeting specific critical algorithms as a mechanism to verify the practicality and effort. Moving the pilot system to production provides validation of the concept of image based inspection as feasible although requiring significant changes in the business processes to gain full efficiencies. The architecture requires integrating diverse components including Windows servers for camera servers, Linux analysis platforms and data integration from mainframe systems. Given scale of the system, the architecture leaned on open source solutions including OpenCV for computer vision. Deep Learning vendor proved the viability of TensorFlow as part of a overall processing flow to verify the serviceability of small rail car components. Tight processing requirements still leaves opens the opportunity for use of AWS or internal Hadoop infrastructure for complex analysis.Developing the full library of algorithms for the check conditions will takes extensive investment. The initiative is current reevaluating the cost benefit analysis for extending the services.

Managed requirements, vendor selection, and implementation planning for a Fleet Telematics solution. System solution was driven by new DOT requirements for GPS tracking of heavy Trucks and Hours of Service for fleet of trucks across Eastern US and Canada. Solution used Software As A Service to provide certified solution of certified truck hardware and driver tracking systems. Actively worked with the vendor to customize new rules such as territory based exceptions that optimized the solution for the railroad. Solution provide Telematics services to the truck and provided CSX with vehicle-based hot-spots to support multiple corporate systems.

Managed development and maintenance of a Linear Asset Management system. The Integrated Track Inspection System is an industry leading system that provides trackmen with a tool that directs inspections of 30,000 miles of linear assets according to complex regulatory schedule rules. The solution implemented a combination of C# user interface, Web Services, and SAP data services to provide inspectors with context sensitive options for track inspection.Managed the development of a new computer maintenance management system for linear track and point rail assets. Track Works built on existing services infrastructure to provide a tablet based mobile application with extensive business rules engine for management and scheduling of linear asset repairs.