Design and Appraisal of a Commercial Multimodal Tomography System

H. I. Schlaberg1, C. Qiu2, F. J. W. Podd3, X. Jia1, K. Primrose2 and G. Bolton2

1 University of Leeds, Institute of Particle Science & Engineering, Leeds, LS2 9JT, UK, Email: H.I.Schlaberg@leeds.ac.uk

2 Industrial Tomography Systems Ltd 47 Newton Street, Manchester, M1 1FT, UK

3 University of Bradford, School of Engineering, Design and Technology, Bradford, BD7 1DP, UK

ABSTRACT

Conventional tomography systems rely on a single measurement modality to extract information from an industrial process. While for many applications a single modality will provide sufficient characterisation of the process, the inclusion of further measurement methods provides additional and complementary information that can be essential. Employing separate instruments for each modality raises issues of synchronisation, modality cross-talk and interference. It also presents difficulties for the data fusion and interpretation of the separate data sets. Centrally controlling the data acquisition from multiple sensor types and closely integrating the datasets within the same instrument, eliminates many of these issues, resulting in combined data of higher integrity.

Research work over several years at several institutions established the fundamental basis for this objective. With further investment and drive by Industrial Tomography Systems Ltd, this work has gone from an university prototype to a commercial multimodality instrument, the ITS M3000. The addition of the ubiquitous USB 2.0 interface at the commercialisation stage, in place of a proprietary PCI interface, adds to the simplification and easy deployment of the instrument. At the same time, this demonstrates that the modular architecture of the system hardware and software allows for rapid customisations and upgrades.

The M3000 system comprises three tomographic modalities: electrical resistance, electrical capacitance and ultrasound tomography. This paper describes how the initial difficulties of reliably combining multiple measurement modalities were overcome by tightly integrating the hardware and synchronising the data acquisition within a single instrument. An important design consideration was that of synchronised acquisition and time-stamping of the data, thus permitting simultaneous views of different process aspects and facilitating data fusion.

A single graphical user interface allows control of the system and performs post-processing and reconstruction tasks. This permits new data processing configurations to be performed without the need for writing code. Results from the integrated multimodal measurements and reconstructed images are presented.