Project: High-efficiency Computation Platform for Massive Virtual Metrology Data Management

European manufacturing industry is focusing its efforts, following 7FP initiatives, to improve the development of new, sustainable and innovative production technologies. Most of these technologies are based on process automation and virtualisation._x000D__x000D_Optical technology is playing a key role in the factories of the future as it delivers high-speed data acquisition and high-accuracy for virtualisation of components for quality and security controls. Most recently developed optical sensors optimize the data acquisition over very large parts and components (aircraft wings and fuselage, windmill blades, etc.). Consequently, image, multimedia and raw data obtained in the different manufacturing processes by these techniques is nowadays immense and offers a lot of possibilities and solutions. However, while the 3D data acquisition process has been remarkably improved through optimized optical sensing technology, the efficient processing of such vast amount of 3D information is still very limited._x000D__x000D_Thus, the challenge now lies on providing a high-efficiency computation platform to make possible the fast processing of such large data sets of image and information within reasonable timescales for intelligence extraction and decision support during the manufacturing processes. This perception is reinforced by the fact that within the wind generators and aeronautical components manufacturing sectors there is a current need to look for advanced solutions in terms of speeding up the processing of large point clouds obtained by optical sensors._x000D__x000D_Next generation windmills may consider blades as large as 100 meters, and A380 or A350 aircrafts will increase the dimensions of their wings and tails in a near future. This implies that physical manipulation of such large parts becomes more complex, expensive and represents a serious risk to health and safety of workers. Therefore, manufacturing companies are starting to consider performing such quality control operations over ultra-large parts in a virtual way as a mean to improve competitiveness. So, accuracy in 3D object virtual processing can be achieved while risks and costs of physical part manipulation can be mitigated._x000D__x000D_The CLOUDS project aims at developing a new high-performance computing platform for extremely large image/multimedia data sets processing with the objective of speeding up manufacturing solutions based on virtual data treatment. It will allow that unprecedented 3D virtual manufacturing processes can be executed very fast. Basically, the speed of the data processing will be increased because the large amounts of data shall be divided, treated simultaneous in different computers and put back together again._x000D__x000D_The research and technological development goal is to improve and combine existing state-of-the-art technologies in order to build an efficient multi-core parallel programming server, installed at the project leader facilities, optimized for performance and productivity of SMEs and suitable for critical large scale data intensive applications. This server will be accessible by customers via an application or system that accesses a remote service on another computer system through a network. This is an important innovative aspect of CLOUDS: manufacturers will not need to invest in a very expensive computer system to process their acquired data. CLOUDS application software will allow them to choose the kind of processing they need in an easy and friendly user interface, and it will send the information compressed to the server that will carry out the selected tasks and send back to the customer the results for their exploitation._x000D__x000D_The use of multi-core computing technology and cloud computing solutions will be essential on one hand to ensure that computing intensive applications can be executed optimally over multiple CPUs and GPUs. On the other hand in order to optimize and distribute transparently over multiple computers the heavy computational work required for parametric analysis and AI system training for every particular application._x000D_TRIMEK, Spanish SME leading provider of virtual metrology solutions for quality control, will be the project coordinator due to its knowledge of the manufacturing sector, principal market were the project results will be addressed. TRIMEK will subcontract the services of Innovalia Association, an independent private Associated Research Lab, organised into three different units: ICT; Nanotechnology and Advanced Metrology solutions; and Manufacturing and Knowledge Management._x000D__x000D_ActiveEon is a Professional Open Source Software Company originated from INRIA (French public-sector scientific and technological institute under the dual authority of the Ministry of Research and the Ministry of Industry) that co-develops innovative software: ProActive Parallel Suite, a complete framework providing a Java Parallel Programming library seamlessly integrated with Scheduling and Resource Management tool.

Acronym CLOUDS (Reference Number: 6658)
Duration 01/06/2011 - 31/05/2013
Project Topic The CLOUDS project aims at developing a new high-performance computing platform for extremely large image/multimedia data sets processing with the objective of speeding up manufacturing solutions based on virtual data treatment.
Network Eurostars
Call Eurostars Cut-Off 6

Project partner

Number Name Role Country
2 ActiveEon SAS Partner France
2 TRIMEK S.A. Coordinator Spain