Manufacturing Analytics

Manufacturing is not all about the predictable and the controllable – factory systems, supply chains, product lifecycle networks are partly emergent, partly designed
complex systems. The explosion of data, digitalisation and artificial intelligence in manufacturing and supply chains helps us both understand these complex systems
and gives us tools to be prepared for the unexpected.


The Manufacturing Analytics Group (MAG), part of the Distributed Information and Automation Lab (DIAL), studies emerging artificial intelligence technology, and
nature-inspired and agent-based computing techniques to develop novel tools and methods for understanding and handling emergent outcomes in industrial systems.

 

Supply network analytics


We create methods to discover hidden patterns in data that yield useful insights for improving supply chain operations. These insights can be used to forecast deliveries, disruptions, find out hidden information, predict quality of goods and even estimate the best price for procurement negotiations. Once the system state has been predicted, then autonomous algorithms can control daily low-level operations to nudge supply chain systems to a more desired state. Current activities include:

  • Autonomous supply chains using agent-based systems
  • Predicting “hidden dependencies” in supply networks
  • Predicting disruptions in supply networks
  • Optimising supply network reliability
  • Study and detection of self-organisation and emergence in supply networks

 

Factory analytics

 

Our research aims to create AI systems that process and analyse data generated within a factory or a production system in order to optimise quality, throughput and utilisation. For example, data from metrology equipment could be used to diagnose likely trends in quality and determine where appropriate interventions should be placed. Current activities include:

  • Quality prediction using IoT data
  • Predicting fresh produce demand and optimising yield
  • Self-organisation and emergence in factory operations

 

Analytics and decision-making in self-organising industrial systems


New cyber-physical systems in manufacturing result in autonomous systems where decisions are made by a network of interacting agents. Examples include distributed planning of production, supply chain deliveries and services. This theme focuses on developing analytics solutions to enable such systems to make optimal, on-time decisions using data.

 

  • Collaborative learning for improving small scale production output
  • Coordinating procurement decisions in self-serving assets

For further information please contact:

Dr Alexandra Brintrup

T: +44 (0)1223 764615

E: ab702@cam.ac.uk

Share This