Impact case study

Ewan is a Cohort 3 doctoral researcher with the EPSRC/NERC CDT in Offshore Wind Energy and the Environment, hosted by the University of Sheffield in partnership with Siemens Gamesa Renewable Energy.

Ewan’s research focus is on building models updated by live data to better understand the manufacturing process for offshore wind turbine blades.

Supervisors:

Prof Peter Osborne, Professor of Advanced Manufacturing, University of Sheffield Advanced Manufacturing Research Centre 

Prof James Gilbert, Director of the Energy and Environment Institute, University of Hull

Prof David Wagg, Professor of Nonlinear Dynamics, University of Sheffield

Dr Steven Balding, Siemens Gamesa Renewable Energy

To begin the project, we needed to establish what the biggest inefficiency in the process is. To do this, structured interviews were carried out with staff from the Siemens Gamesa factory in Hull. The results from these interviews showed that waiting, wasted time due to a slowed or stopped process, was the most reported inefficiency. To better understand how this waiting was occurring and how to mitigate it, a factory simulation that captures how the process occurs was proposed.

To build this model, an agent-based modelling approach is being used. To ensure that the final factory model can be trusted, each part of the process needs to be individually created and checked in isolation. To do this, a part of the process is modelled in mesa, an agent-based modelling package for python, and a mathematical model for how that part of the process should behave is created. The simulation model is then checked against the mathematical model to ensure that the correct behaviour is being carried out.

Following the creation of the final factory model, the accuracy must be checked and the model updated accordingly. To do this, historical production data from Siemens Gamesa is being used to compare the model outputs to real scenarios. The model will be updated accordingly to reflect the real production process. Once the model can be shown to behave as expected and reflect real life scenarios, it can be used to test the impact of different constraints on the production process to build a framework of how the Siemens factory behaves under changing conditions.