Hewland Harnessing the Power of Data & AI

---

Hewland has recognised the increasing influence that data analytics and AI are having on society and their importance for the future of the engineering industry.  We are on a journey to harness the power of these tools to improve processes, support sustainability initiatives, inform engineering led decisions and meet the long-term expectations of our customers.

For our first foray into data analytics and AI, we harnessed the vast amount of data generated by our manufacturing facility; our primary sources being energy usage and real time workflow data.  An initiative by Hewland’s sustainability team installed monitoring equipment on the factory floor to record the energy usage of the machines.  In parallel, our Smart Factory tool allows for processes and workflows to be continuously monitored throughout the facility. 

Though separately both these tools provide valuable information, it is when they are combined that a greater understanding of our production process can be gained.  An in-house developed python code links the real time data from the two sources and allows key metrics, such as manufacturing time, cost per operation and machine utilisation, to be quickly calculated and reported.  Ready access to these key metrics allows for process improvements to be identified and the impact of their implementation to be quantified.

A further benefit of combining and summarising these two data sources is that it provides a wealth of training and validation data from which machine learning models can be created to forecast key information for new Hewland products, such as total production cost and manufacturing time.   Though the creation of accurate forecasting models is the primary goal of the proof-of-concept phase, the knowledge gained has provided a springboard for greater application and adoption of AI throughout the business. 

During the AI application phase, the following initial activities were identified:

• Taking on the recommendations of industry experts, a working group was formed to identify opportunities to utilise AI throughout the business.  Particular focus was given to the predictive maintenance of our machines and quality analytics.  It was also the responsibility of the working group to govern the usage of AI to ensure it is responsibly utilised.
• To complement Hewland’s new state of the art test facility, improved data analytics have been applied to the test results using the lessons learnt from the proof-of-concept phase.  One aim includes the use of the real time data to perform condition monitoring; for example, in the case of loss of lubrication testing this would be to identify from the many sources of test data when the criteria has been met to stop the test.
• Hewland has been quick to adopt and implement new CAE methods into our design workflows, supplementing the expertise of our engineering team.  Reduced order modelling and deep learning methods have the capability to both accelerate the design process and broaden the range of conditions considered in the analysis.  Structured studies will be performed to utilise and integrate these new approaches.

The focus of all these activities is to improve existing processes and, through the competence gained, for Hewland to be better placed to support the future needs of our customers.  Within the transmission sector, increasing focus is being given to digital twins and sustainability initiatives to ensure circularity and life cycle thinking is at the core of transmission design.  For specialist transmissions operating in an environment where down time due to component failure is unacceptable, either from a cost or operational perspective, digital twin technology provides a means to continuously monitor each individual transmission in service and calculate the damage they have accrued. 

The twin can then automatically notify when preventative maintenance is required.  As highlighted in the figure, completion of the initial activities in the application phase will enable Hewland to provide this service for our bespoke transmissions.  Finally, all the techniques mentioned above can facilitate continuous improvements of our products, while reducing carbon footprint.