EPSRC Fellowship in Manufacturing: Collaborative Metrology Systems for High Value Manufacturing

To support the development of challenging, difficult to manufacture products, increased reliance is placed on techniques to allow accurate dimensional measurement of parts and components. New measurement systems are needed that provide data quickly with higher levels of accuracy and precision than is currently possible. Currently high accuracy measurements are made using dedicated expensive instrumentation in temperature controlled labs. The wide range of measurement challenges mean there is no single instrument available to suit all needs. In fact, the range of lab based instrument systems required to meet the measurement needs of industry continues to grow. It includes techniques ranging from contact measurements made using a mechanical probe, to non-contact measurements which use light, lasers, or X-ray based measurement methods. The main drawback of these systems is that they are usually slow to set-up, and significant time is required to take measurements. This means that although they are very accurate they are less useful for the control and improvement of challenging manufacturing processes, where data must be collected and analysed quickly.

Improved measurement systems are required which provide higher speed measurements, at lower cost, without compromising accuracy. Currently two approaches address this need. One approach uses on machine sensors to provide high-speed measurements, while the other approach is to position instruments closer to the manufacturing environment to reduce the time required to transfer work to the measurement lab. Both approaches have obvious benefits as they provide faster data; however, they are also less accurate as a result of the unwanted disturbances experienced on the factory floor. These limitations result in a trade-off: the user can either have high accuracy, or high speed measurement, but not both at once.

The research undertaken within this Fellowship will develop a new way of collecting and effectively processing critical measurement data. Instead of a reliance on high accuracy instruments, this approach will provide a new way of thinking with respect to how measurement systems are designed and implemented. The goal will be to allow different types of lower accuracy data to be combined in a beneficial way. For example, computer simulations of a machine, product, and process will be combined with sensors that monitor environmental conditions. In addition sensors used to take high speed measurements of parts during the manufacturing process itself will be used. Through a collaborative process these data will be combined to provide fast high quality data. To verify and further improve the system a small quantity of accurate feedback data from high accuracy instruments in temperature controlled labs will be used. In effect the approach will be to combine slow accurate data, with fast less reliable data, to produce enhanced accuracy fast measurements.

For example, if a batch of high precision components must be produced, the components must also have their geometry verified and corrected if required. On machine sensors may be used to verify geometry, but accuracy is limited due to environmental effects such as temperature and humidity. To compensate for these errors a collaborative measurement system will initially make measurements using both on-machine sensors as well as off-machine lab instruments. It will blend these data sets in addition to data from on-machine environmental monitoring sensors, and computer simulations to correct for errors and therefor enhance the accuracy of the measurements. The system will automatically adapt to changing environmental conditions by triggering the need for more lab-based data which will allow an improved error correction to be made. In this way the system will adapt and optimise the measurement process to suit the current manufacturing conditions.

Metrology plays a critical role in underpinning the development and optimisation of all new manufacturing processes and technologies. Therefore this work will be of significant benefit to industry interested in the development, design or manufacture of difficult to produce high-value products, such as micro devices, sensor packaging, hard to machine aerospace components, or mass customised medical devices. These products all require high tolerance to be maintained, required high quality assurance, and present a serious challenge to manufactures. In addition for the development and successful application of emerging technologies such as additive manufacturing for example, metrology will pay a pivotal role in effective process optimisation and control. The fundamental metrology know-how created will allow more responsive and more accurate metrology systems to support scientific development, engineering design, and manufacturing.

During the development of this proposal a central recurring theme became apparent following discussions with the selected industrial partners. There was general agreement that the collaborative use of data will necessarily become an integral part of future manufacturing. WLR Precision Engineering commented that as an SME their future business model is based on the production of difficult to make products that cannot easily be sourced from low cost-base countries. They also commented on the rising trend for their customers to ask for increased quality assurance data from multiple measurement sources. Rolls-Royce stated that while they have examples of combining multiple datasets to enhance measurements, they lack a clear methodology for how to implement this efficiently. They also highlighted the significant lack in techniques that allow useful process control information to be extracted from the large quantities of data that can be generated in a modern manufacturing environment.

As such the general comments from industry experts at Rolls-Royce and WLR Precision Engineering were that developing a better understanding of collaborative metrology will be critical for future success in a competitive global market. It was also acknowledged that currently there is significant inertia preventing industry from moving in this direction. Notably a lack of fundamental knowledge, concerns over feasibility, and practical issues such as data security were typical concerns. This proposal will address these issues by providing fundamental knowledge with the foundation for on-going research, combined with practical demonstration of the possibilities and benefits of the proposed techniques. This fundamental knowledge will form the basis for future metrology standards related to the analysis of large data sets, and methods for combining and analysing multiple datasets.