Artificial Intelligence in Engineering Management

Project Description

Applications are invited for a fully-funded full-time PhD studentship (UK/EU applicants) ’Artificial Intelligence in Engineering Management.’ The project is in collaboration with Dyer Engineering Ltd: https://www.smarterstrongertogether.com

The purpose of this project is to advance the application of AI in Engineering management, in particular, improving the quotation feedback process of the industrial partner Dyer Engineering Ltd. This may include the following areas of investigation:

• Automatically analyse engineering CAD as well as paper-based drawings (recognise features).
• Analyse project-related data/systems to generate the initial quotation (maybe Excel-based, maybe database or unstructured data).
• Analyse Enterprise Resource Planning software to determine how individual features were realised (learning from past projects).
• Compare As-Is (ERP data) and To-Be (initial quotation).
• Develop a learning system to improve the accuracy of future quotations.

The research of this AI project will be focused around the need to create a ’feedback loop’ of continual improvement of enquiry prioritisation, quotation accuracy (plan vs actual), commercial base costing verification and automated review, shop floor communication and production process optimisation. In other words, the proposed project aims at investigating the quotation process to neither reduce profit nor lose business due to unrealistic quotations.

This is likely to implement a recurrent neural network (RNN) since the length and structure of the quotation data may vary enormously for different contracts. The analysis of quotations can be seen as a big challenge for an RNN. First of all, the RNN has to understand the natural language used to encode the information in the quotation. This can be achieved with embedding layers in RNN. The second and most crucial challenge for an AI is to decode the encoded information in the quotations and encode them into the internal states of the neural network. Here is where the full power of RNNs can be exploited. What described so far is the "encoder" of the RNN; the other part of the RNN is the "decoder" that in this context provides useful suggestions to improve quotations.

Dyer Engineering is a rapidly expanding business which manufactures metal components and structures and delivers related services. The business operates across a diverse range of markets, with the ability to manufacture parts you can pick up by the handful, through to large-scale structures operating in harsh subsea environments. They also provide support services to customers ranging from paper mills to aerospace companies: https://www.smarterstrongertogether.com

Funding and Application Process

The project will take the form of an industrially supported studentship with the student based at Durham University as well as working with Dyer Engineering during the 3-year fully-funded study programme.

Entry Requirements: Applicants should hold at least a 2:1 honours degree or equivalent in Engineering, Computer Sciences, Applied Mathematics or a related subject.

Eligibility: This studentship is supported through the Intensive Industrial Innovation Programme, part-funded by the European Regional Development Fund. The studentship covers full Home fees and a tax-free stipend at the EPSRC rate. Applicants from outside the UK/EU are not eligible for this award unless they are able to self-fund the difference between UK/EU domestic fees and international fees (https://www.dur.ac.uk/study/pg/finance/costs/)

Start date: 01/10/2020
Location: Durham University Engineering Department and Dyer Engineering Ltd

Contact: Informal inquiries can be made to either Dr Oliver Vogt or Dr Stefano Giani

How to apply: To make an application, please visit the University applications page at https://www.dur.ac.uk/study/pg/apply/. When making an application, please state the project title "Artificial Intelligence in Engineering Management".

Deadline for application: 31/08/2020.
Funding: £15,285 p.a. based on 2020/21 stipend.