Virtual Reality Tele-collaborative Integrated Manufacturing Environment (VRTIME)

Alan D. Hudson, Brian Dodds, Jim Curtis- Searle; Fred Dech, Pat Banerjee, Andy Banerjee- Industrial Virtual Reality Laboratory

Searle, a Monsanto Company, is a research-based company headquartered in Skokie, Illinois (USA) that develops, manufactures and markets prescription pharmaceuticals and other healthcare solutions worldwide. VRTIME is a tele-immersive system targeted at industrial process design, operation, and control. Searle's operations in Britain, United States and Russia will be used for long distance TI case studies. In the industrial process planning phase, VRTIME will be used for facilities design, regulatory review and operational review. This is a particularly good application for TI because these reviews often involve geographically separated participants: architect, engineers, consultants, external and internal reviewers, federal, state and local planning officials, and environmental interest groups.

In addition VRTIME will be used to train new staff and to retrain experienced staff working in new or reconfigured facilities. The project will focus on training in two areas: shop-floor management, and quality control of a manufacturing system. At the shop-floor management level, the training activities will assist staff to achieve optimum performance before they operate real processes. For training at the quality control level, a simulated, virtual manufacturing system will be developed. On the top of this virtual manufacturing system, a quality control system based on the Unit-based quality control methodology will be built to determine the quality prediction and the confidence of the prediction from simulated process data. The prediction engine will have been modeled using data-mining algorithms that are applied to real process data (gathered from measuring instruments in the factory) and quality data (determined when a final product has been manufactured). Both the simulated, virtual manufacturing system and the quality information will be presented to operators who make decisions on the fate (e.g. proceed, scrap, rework) of production units.

The tele-immersive nature of these training sessions means that they can be guided by remote teachers and undertaken before the new plant is operational, enabling a rapid start-up phase when the plant becomes operational. We are currently applying some of these techniques to a new plant being built in Moscow, Russia. The use of
LIMBO in this training system has allowed us to jump start development by allowing us to quickly build a tele-immersive environment by importing existing three-dimensional models of the plant's production units.