Imperial College London and Three Media implement “near” real time platform optimisation
Research project results in significant performance improvements, simulation accuracy, and AI-driven optimisation for any workflow or business process
Manchester, UK – 12 November 2019 – Three Media, media specialist and product provider of XEN:Pipeline, today announced the completion of an 18-month project in conjunction with Imperial College London. The goal of the project was to make media supply chain processing more affordable by identifying ways to reduce costs and maximise workflow, process and resource efficiencies.
The project, called Digital Media Data Analytics (DMDA), was initiated and led by Three Media in conjunction with an Imperial College London team headed by Performance Engineering expert Dr Anthony Field. The combined team applied its expert knowledge of data analytics and optimisation techniques to Three Media’s XEN:Pipeline development programme.
XEN:Pipeline enables users not only to evaluate and optimise existing workflows and business processes in near real time, but also to model new and changing virtual revenue streams either as stand-alone or alongside the existing. It also provides opportunities to identify, model, and prove new and innovative services and chargeable functions. XEN:Pipeline is a highly configurable, scalable, storage agnostic and performant system that can operate across multiple company and supplier locations and departments, identifying and resolving often-hidden bottlenecks, pinch points and under-utilisation from existing workflows and processes.
Three Media Partner and CTO Craig Bury said, “We know from talking to our clients that their biggest challenge at the moment is to drive down their operating costs but with the knowledge that they can guarantee to provide the same level of service. Based on this, the aim of this research project was therefore two-fold. Firstly, to increase the performance of XEN:Pipeline so it can optimise in near real time with constant SLA evaluation and secondly to minimise media supply chain costs to make processing much more affordable, which in turn can enable more content to be exploited.
“Simulation modelling provides ways to reduce costs and increase efficiencies by identifying how revenue can be maximised, ensuring technical and human resources are scheduled only when needed. Human resources can also be redistributed or deployed in more logical shifts to engage in more productive roles. With XEN:Pipeline and its enhanced capabilities, the frustration of wondering why staff are sitting around waiting for something to do or why additional transcoders or storage don’t improve throughout will no longer be an issue.”
Imperial College’s Dr Field added, “The XEN:Pipeline platform was already an excellent tool with extraordinary modelling and system optimisation functionality. However, by applying our expertise to the DMDA project, we have been able to accelerate the performance of the simulation component of the XEN:pipeline, and simultaneously reduce the volume of data generated by the simulation, by at least an order of magnitude, without loss of fidelity. This has enabled the XEN:Pipeline and other Three Media products to model and optimise much larger and more complex media workflows than was previously possible. Analysis of the simulation output has also enabled us to pinpoint bottlenecks in workflows which subsequently become the targets for optimisation.”
XEN:Pipeline is offered as a Platform as a Service (PaaS) via the cloud or local instances. Its infrastructure supports all existing and foreseeable global media processing platform services. Its design, following the development based on the joint DMDA project, incorporates support for complex data schemas and management of “big data” and associated analytics, artificial intelligence (AI) for optimisation and general machine learning (ML) capabilities. XEN:Pipeline provides SLA, technical, functional, and workflow capabilities that are not available anywhere else in the media industry.