Over the last few decades data has evolved from being well structured and typically housed in single databases and repositories, to the complex “big data” we see all around us. In particular, data is no longer neatly packaged into single silos ready for traditional analysis and analytics. Instead, real-world data is complex, comes in different varieties and formats (pictures, documents, video, images, transactional data, time series data, sensor data, maps etc), suffers from serious data quality issues (often incomplete, inaccurate out-of-date and even biased) and is regularly spread across multiple systems, databases and silos, both on-premises and cloud based.
At AWL we have been evolving and adapting to meet these new data challenges head-on. Our Data Science and Data Intelligence team are leading, researching and developing cutting-edge data solutions and capabilities, leveraging the latest advances in data management, artificial intelligence, machine learning, and data intelligence.
We believe that as data volumes continue to grow exponentially, as data complexity increases and with new data varieties still emerging, it is imperative that we deliver “data led” solutions and services, rather than being “data driven” or “data pushed”.
Artificial Intelligence and Machine Learning
Whether it is standard supervised (regression, Support Vector Machines, decision trees and random forests) or unsupervised (clustering, dimensionality reduction, association rule mining) machine learning methods, or the latest advances in deep learning and natural language processing, data and AI are at the heart of our products and services.
As a trusted partner to the UK MoD and other key defence industry partners, we are working with them on a range of AI and data projects.
For example, in conjunction with the MoD we recently developed an AI model to predict obsolescence risk of equipment aboard Royal Navy vessels. The outcome of our solution was an increase in accuracy of 75%, allowing the client to better predict the obsolescence status of key parts aboard Navy vessels.
In addition, last year we commenced an R&D phase to apply advanced computer vision techniques to identify parts and items of machinery from technical drawings with the aim of developing a solution capable of assisting with NATO codification tasks, as well as extracting information from Illustrated Parts Catalogues.

Graph Analytics and Capabilities
At AWL we are partnering with some of the world’s leading graph database and graph visualisation companies to develop new cutting-edge solutions for a range of use-cases.
For instance, we are currently developing a graph based system for Master Data Management, working across disparate silos to reveal the relationships hidden in master data as well as deliver single sources of truth.
We are also using graphs to understand complex Bills of Materials, to understand the dependencies between components and equipment across military assets, as well as understanding how spares and parts can be sourced in complex supply chain networks.
We are also researching how Natural Language Processing (“NLP”) techniques, OWL based ontologies and Resource Description Framework (“RDF”) graphs can be used to build insights between entities and relationships from technical and other documents.
