What is Data Science?

When people say ‘Data Science’, it’s sometimes just thrown into conversations as a buzzword used to describe everything from Excel spreadsheets to deep neural networks. Personally, I would define Data Science as using scientific methods and structures to extract useful insight and knowledge from structured and unstructured data by applying statistical analyses. This includes Machine Learning, traditional statistical modelling or more advanced deep learning techniques. It can be applied to structured data held in databases, but increasingly to unstructured data such as text, images, audio and videos as well.

The value data brings to a business

Organisations are often sitting on vast amounts of data that could help them become more efficient and identify risks and opportunities more easily. The most successful businesses of the last decade are those that have capitalised on this data to help them improve internally, but also to adapt to customer needs. Compare Netflix to Blockbuster or Asos to Debenhams. Data is at the heart of every successful business and Data Science allows you to extract the insight from that raw data.

Techmodal’s Data Science capability team

As a team, we have experience in exploratory data analysis, machine learning, forecasting, natural language processing and computer vision projects from proof of concept through to delivery. The team is made up of a growing number of extremely talented data scientists ranging from recent graduates to those with extensive commercial and technical experience. They are all passionate about the value data science can bring and are driven to develop technical solutions to solve business problems.

Examples of Techmodal’s transformative Data Science capability in action

The Techmodal Data Science team is currently delivering a wide range of solutions and products to the UK Ministry of Defence and industry; tackling challenging problems with zest. Amongst these exciting solutions, I would like to highlight the following recent examples now in use with our defence customers and which were delivered within a tight schedule:

HUMS is Techmodal’s predictive maintenance model which ingests streaming Health and Usage Monitoring Systems (HUMS) data from military vehicles and is utilised as a key enabler to ensure the readiness and availability of the platforms. This powerful data management and dashboarding tool was delivered, utilising platform data from hundreds of sensors, to underpin prognostics and Through-Life Support decision-making, being expanded with Machine Learning. Through analysing this data, the user is able to detect deviations from normal operating levels and pre-empt critical failures.  This allows those on board, as well as the maintenance and management teams back at HQ, to clearly observe and accurately predict the likely failures and address them with preventative action, well ahead of time.

A second example is the Techmodal: PE product which applies an AI-driven prediction engine to ensure capability sustainment of the Royal Navy’s Carrier Strike Group.  This Techmodal product enables the user to accurately predict future defects across platforms as well as model the current state of equipment and materiel.  This is integrated within a user-friendly dashboard to give decision-makers the information they need to proactively manage the platforms.  This means identifying at-risk systems, providing context on the impact of a failure in terms of capability loss, and providing all the data required to formulate a solution. 

In addition to these solutions in service today, this third example reflects an aspect of our future technology roadmap with the design of systems and models that will deliver control rooms for the next generation of submarines.  The Techmodal Data Science team has developed an ontological model that can sit seamlessly between human operators and computer systems.  This ontology is hosted on a graph database and uses the latest approaches from academia and industry to future-proof the development. 

About the author

Tom Fowler is the Data Science capability lead within Techmodal. He provides team leadership and direction, supporting with hands-on delivery of projects, and also provides expert technical advice to the sales team.

Tom began his career working for the Police, breaking down barriers for information sharing, and developing Machine Learning models to support vulnerable people across Bristol. This work was widely recognised and was recently profiled in the New York Times. Following that, Tom moved into consulting and has worked across the health, local government and defence sectors.