Artificial intelligence and data science are required as cross-cutting capabilities across a range of industrial and commercial sectors, reflective of the applied nature of the majority of our research and our strong industrial links.

Our expertise and application

Cranfield academics have developed skills and expertise is a range of discipline specific methods, tools and techniques, such as;

  • Probabilistic methods for uncertain reasoning including Bayesian network, Hidden Markov model, Particle filter, Decision theory, and Utility theory
  • Logic programming and Automated reasoning
  • Classifiers and statistical learning methods, including Machine learning and non-supervised learning (e.g. surveillance video data, crowdsourcing product design)
  • Data orientated approaches such as neural networks and fuzzy logic including extreme learning, reinforced learning and deep learning (e.g. for air traffic management, classification of damage)
  • Search and optimization including search algorithm, combinatorial optimization, and
    evolutionary computation
  • Extracting, Transforming and Loading processes for effectively transforming raw data for database population e.g. Dimensionality Reduction using PCA
  • System identification methodologies for linear and non-linear systems (Stochastic Subspace
    Identification and non-linear Bayesian methods).
  • Global optimisation algorithms (e.g. evolutionary, nature-inspired, swarm intelligence)
  • Development of human-like inference techniques and learning schemes
  • Natural language processing
  • Comparison to physical models which capture known design and performance characteristics of machines and processes
  • Rule based systems which capture knowledge and standards (popular with customers who don’t like neural networks)
  • Data mining, fusion and visualisation.

We currently offer the following relevant taught courses, although we have a large number of PhD studentships centred in AI and Data Science (most with industrial sponsors) as well as short courses:

In addition, we lead a NERC CDT ‘Data, Risk & Environmental Analytical Methods for Doctoral Training’ aiming to skill and secure the next generation of data scientists and informaticians working in environmental sciences. 30 postdoctoral students are / will benefit from training that combines excellence in risk mitigation science with cutting-edge big data interpretation and allows the students to apply practical big data analytics to challenges faced by industry, academia, NGOs and government. We are also currently exploring the market for a course in financial technology, situated within our School of Management, covering block chain technology, cryptocurrencies, smart contracts, cyber security and utilising big data to profile customers using all available information (including social media).

Our facilities

Our AI, data science and robotics capabilities

Our AI, data science and robotics capabilities are grounded in finding solutions to practical challenges. As such, Cranfield research and teaching is applied to a variety of aerospace, transport, water, energy, defence and security, and manufacturing problems for diagnostic, prognostic, and knowledge management perspectives. For example, our activities use AI to underpin aircraft design, optimisation and trajectory resolutions, supporting safe AI in terms of network computing, developing autonomous systems for drones and self-driving cars, and assuring flow monitoring and control in oil and gas pipelines. There is also a current research cell exploring the broader issues of AI.

Examples of Cranfield data science interests include transforming raw data and data analytics for airline maintenance operations, radar and imaging signals, machine learning for signals and damage classification, identifying materials in waste management, transitioning towards a re-distributed manufacturing model for consumer goods, and structuring knowledge in high value manufacturing and maintenance service provision, managing staff turnover and supporting high reliability equipment.

Our robotics interests lie in engineering and automation although the ethical design and application of robotics, the impact of and skills required for AI, and the experience of the end user of AI-based systems are also a focus. For example, driver behaviour and psychology, human computer interactions in unmanned vehicles and human factors considerations on cognitive psychology. One of our academics, Dr Sarah Fletcher, was also involved in development and release of the first ever standard to guide the ethical design and application of robotics (BS 8611).