Massive Analytic has developed Oscar AP, a unique proprietary precognitive data science and analytics platform.
Persistent surveillance is constant the sensing, detection, classification and analysis of data to provide trustworthy information. An effective military decision making process based on persistent surveillance therefore requires agile real-time analysis of raw data transmitted via numerous different communications networks. Existing techniques for persistent surveillance are manually intensive with considerable demands on highly-trained computer scientists and specialist surveillance analysts. In a time when military manpower is limited, incumbent approaches are therefore too time consuming and a data flow bottleneck within the chain of command arises due to the intensity, complexity and encryption of data, as well as the speed and level of insight required.
The fundamental lack of scalability and performance of existing methods for big data analytics – in terms of bandwidth and insight generation to discover data-led evidence of threats – limits the opportunity to automate a significant portion of the persistent surveillance analysis workload. The capability of current manually intensive persistent surveillance systems to yield the required operational value from big data is therefore fundamentally restricted.
In response to the immediate requirement for a step change in analytical and data visualisation capability for big data across a range of commercial, industrial and defence applications, Massive Analytic has developed Oscar AP, a unique proprietary precognitive data science and analytics platform. Oscar AP includes a range of integrated capabilities including topological analysis of real-time multi-mode data streams, synthetic intelligence, machine learning, predictive analytics and artificial precognition for evidence-based human-machine decision making. Initially targeting automated adaptive control for data-led business processes in the commercial domain, Massive analytic have developed capability for: query- and code-free behaviour tracking; enhanced situational awareness; sensor-driven decision analytics; and deep reinforcement learning to drive anomalous event detection, process optimisation, optimised resource consumption/latency and management of complex autonomous systems. Massive Analytic therefore recognised the potential to make a disruptive impact in the field of defence intelligence through a step change in the level of automation for classification and analysis of persistent surveillance data, based on a convergence of Oscar AP with their additional proprietary technologies under development.
Project goal: Massive Analytic sought CDE support to develop a dedicated big data analytics platform for tactical operational planning/support. The project targeted functionality to combine multiple data sources for ingestion and analysis, outputting prevailing static & dynamic information to predict future states. Massive Analytic aimed to implement advanced autonomous functions to mitigate data deluge effects for surveillance analysts, facilitating the dissemination of critical insight within the chain of command via a novel interactive dashboard. The key outcome of the proposed project was a kernel solution based on machine learning, spectral analysis and artificial precognition, augmented by video precognition, geometric data analysis and multi-stream data synthesis. The project sought to initiate a strong collaborative framework within the Defence Science & Technology Laboratory to validate the potential to predict outcomes through models generated automatically, and enable practical integration into conventional persistent surveillance workflows
TBAT worked closely with Massive Analytic to gather technical and commercial information from relevant staff, and respond effectively to any feedback throughout the process. Meetings in person and remotely were used as appropriate, as well as our own directed activities to actively find and review relevant materials. By gaining a thorough understanding of the technology and it’s wide reaching potential, we were able to structure a proposal that directly addressed the scope of the funding scheme and aligned precisely with the Client’s long term technical and commercial aspirations. In particular, we were able to effectively match the core capabilities of Massive Analytic’s solution with the targeted context and parlance of military intelligence. As well as drafting and submitting the proposal, TBAT assisted with developing the project plan, resource allocation and financial forecasting.
The grant application was enthusiastically supported by the Centre for Defence Enterprise who awarded Massive Analytic the maximum contract value for a Phase 1 development. The project commenced early in 2016 and was successfully delivered on schedule at the end of June 2016, with excellent feedback from CDE / DSTL. The documentation TBAT provided as part of the grant due diligence and application process has been instrumental in enabling the rapid delivery of the project. Moreover, Massive Analytic was able to develop excellent relationships with the other CDE competition winners throughout the six-month project duration, with the support of the DSTL-coordinated networking process. In partnership with these entities, Massive Analytic has submitted further bids for second phase of the project, allowing further development of the technology in line with their overall business strategy.
“The documentation TBAT provided as part of the grant due diligence and application process has been instrumental in enabling the rapid delivery of the project”
Dr George Frangou - Director