Abstract - Managing the glass manufacturing process
The Digital Era of Internet of Things and Industry 4.0 represents an invaluable boost for innovation: recent studies claim that connectivity and interactions among parts, machines and humans will make production systems as much as 30 percent faster and 25 more efficient.
These concepts applied to the glass forming process will evolve from single automated cells (first-level independent closed loops, like gob weight regulation or mould temperature control) to fully integrated, automated facilities. These latest are multiple-level, coordinated control loops, of which the BoX system (jointly developed by Bottero and XPAR Vision) represents the first and still unique example towards a labour free plant.
In such context, the big data analysis recently emerged in the manufacturing world is often presented as the key to optimize production quality, energy savings and equipment service, whereas it can only be considered a pre-condition to the achievement of the final goal: the full automation of the forming process.
No real plant automation can be achieved in a Hollow Glass Plant without a complete hot end forming automation.
Bottero will describe its roadmap aiming to reach this target, starting from BoX approach and evolving in an Industry 4.0 framework.
3D simulations of products, material and processes already used in the engineering phase, will be applied extensively in plant operations. Bottero is already and will more and more leverage this approach with real-time data with the purpose to model the physical world in a virtual environment.
The paper presents Bottero’s vision of the future of the glass manufacturing process: an integrated approach to analyse data, define the optimal process behaviour and support real-time decision making. From an automation perspective, this translates into the need to gather measurements and information from different sources, at both hot and cold end, exploiting the most recent techniques coming from Internet Of Things and Machine Learning analysis to build a predictive end-to-end model of the production process, to be put under automatic control and anticipate corrective actions.