The Aerospace Technology Institute has revealed the findings of a study into the current state, opportunities and challenges of big data analytics in the UK aerospace sector.
Big data involves using advanced technologies and processes such as algorithms, machine learning and data mining to extract value and inform decisions.
Today, aerospace companies collect vast – and rapidly increasing – quantities of data. However, data by itself is relatively worthless. Too many businesses, for example, have admitted to collecting data for data’s sake.
Big data analytics helps to generate valuable insights which are then being used to improve decision making across the value chain, from measuring product performance to assessing the impact of weather on the supply chain and how it might affect downstream operations.
Big data challenges
Before analysing data, its fundamental architecture needs to be addressed. Data configurations are not always directly transferrable and legacy systems may no longer be supported. Translating input data into value can be difficult and requires skills or roles not traditionally associated with aerospace – such as data engineers or data scientists.
Utilising data analytics to enhance product services can also be restricted by the ownership of data, something which may be impacted by the arrival of the EU’s General Data Protection Regulation (GDPR) which comes into effect in less than 12 months.
Big data opportunities
Relevant, accurate and insightful data imparts knowledge, facilitates decision-making in real time, and enables improvements across all stages of the product lifecycle.
Opportunities exist where data from real conditions can generate a virtual representation, analyse alternative scenarios quickly, and improve awareness of multi-disciplinary design decisions. For example, adaptive machining can utilise big data analytics to adapt manufacturing based on a range of input criteria, including environmental conditions, component attributes or tool wear.
In the long term, big data and machine learning can be used to react to events where new requests are dynamically scheduled to support real-time support and services to customers.