Gestamp – XR Emulator: Advanced Production Process Visualisation and Interaction in Real Time
Gestamp
Sector: Industry
Business Case
The interactive visualisation of very large models requires prior optimisation processes to obtain complete scenes that can be visualised with a high degree of performance (rendering speed) and model quality (photo realism). Generating a complete plant model connected to reality through IoT solutions.
Objectives
Develop a 3D environment that allows users to have a workspace where they can visualise 3D models of a complete factory on PC, Web, tablet or mobile, with a fluid visualisation. Make the virtual elements of the plant as realistic as possible in terms of geometry, materials and lighting. Connect each of these elements with real plant data with IoT solutions to monitor them in real time.
Use case
The factory’s 3D models are a collection of CAD and polygonal models from different suppliers. There is a need to generate VR-compatible versions that are lighter in weight. Computational geometry techniques are used to: a) simplify the geometry, b) eliminate non-visible elements, c) replace geometries with simpler ones. Interactive mechanisms (cinematics and animations) are added, as well as photorealistic materials that provide the final scene with the required realism. Finally, the factory’s active elements receive real-time information on their status.
Infrastructure
Edge Cloud On premise
Technology
Machine learning and deep learning Computer vision
Data
Modelos 3D: Modelos CAD y modelos poligonales de alta complejidad, heterogéneos y no orientados a la visualización. Cinemática, animaciones y materiales de los elementos 3D. Imágenes reales de la planta real, datos de planta.
Resources
1 project manager, 2 developers/programmers and 2 experts in 3D modelling, VR/XR and usability.
Difficulties and learning
The main difficulty comes from the high variability in the 3D models. It is necessary to process and simplify the models individually taking into account their role and importance. From a scalability perspective, it is essential to delve into techniques that can be automated, either through complete unattended processes or with assistance and review by 3D modelers.
KPIs (business impact and metrics of the model)
Visualisation: 3D plant rendering speed with different configurations. The aim is to obtain 30 fps in a stable manner. IoT: Connecting and processing data in RT to visualise results and kinematics.