M4I4, Mathematics for Industry 4.0

Abstract

By definition, the digital factory represents a network of digital models and methods of simulation and 3D visualisation for the holistic planning, realisation, control and ongoing improvement of all factory processes related to a specific product. In the last five or ten years all industrialised countries have launched initiatives to realise this vision, sometimes also referred to as Industry 4.0 (in Europe) or Smart Manufacturing (US). In my presentation I will use this example to highlight chances and pitfalls in the collaboration between mathematics, engineering, and industry. I will present case studies showing how mathematics will be able to contribute towards digital manufacturing and close with some remarks about the perspective of collaboration between math and industry in Europe.

Abstract

The continuous and outstanding advances of measurement technologies have enabled the collection and storage of high-resolution data which can often be modeled as smooth functions (e.g., curves or surfaces). This kind of data are at the basis of functional data analysis (FDA) which is a well- known lively and expanding research area of modern statistics. In FDA, the classical concept for scalar or multivariate random variable is indeed replaced by the concept of functional random variable. Consequently, in FDA the typical data set is not made of numbers or Euclidean vectors but a collection of functions embedded in a suitable separable functional Hilbert space meant to formalize application-specific relations between sample units. Recent applications of FDA techniques in different and many fields of science are countless. Nevertheless, very few business and industrial applications can be found, thus pointing out the existence of an unexploited potential of this type of techniques in these two fields. With respect to this discrepancy, after a gentle introduction to FDA, this talk will showcase some recent business and industrial applications in which state-of- the-art FDA techniques are fruitfully used.

Abstract

Miningful Studio specializes in extracting and processing hidden information in the complexity of data, transforming them into decisions, making predictions more accurately, and responding readily to the needs of the market. By combining data analysis, algorithms, mathematical modelling, and a team of data scientists, Miningful Studio transforms complex data into meaningful results, communicative visualizations and effective stories. We operate both in business and academic settings. We specialize in the following areas: digital publishing, cultural heritage, precision agriculture, healthcare, and pharmaceuticals. A couple of case studies will be used to show how data science and practical experience are combined to provide data products.

Other Authors

Roxana Istrate, Costas Bekas

Abstract

Neural networks have proven to solve effectively very difficult problems such as object recognition and classification, speech recognition, automatic translation, text generation, up to gameplay. However, the design of new neural network architectures is still considered a challenging task, driven by an empirical process based on human decisions and manual parameters tuning. Even the most experienced architect would take order of months to design a successful architecture for a new industrial-specific dataset.
Indeed, despite the tremendous success of deep learning, there is still a very limited understanding on the underlying mechanism that allows a network to learn. At IBM Research, we realize that a major leap in neural network application on real industrial problems can occur only by unveiling and understanding those mechanisms. Doing that, we would not only be able to train, but also to design new networks automatically, leading to a by far more scalable and widely applicable process. In this talk we briefly present our current research advancements in this direction, and discuss how potentially mathematicians can help us achieving this ambitious goal.

Abstract

This talk will provide my personal viewpoint about the following issues:
• Is Industry 4.0 an actual revolution? • Does IoT really make things smarter? • Are big data really big? • Which is the role of mathematics and of mathematicians in all this?
In order to discuss such viewpoint, I will utilize examples and results concerning the application of mathematical models and methods in physics and physiology.

Abstract

Manufacturing industry is under a rapid development for both production processes and products, with the making of high flexible and customizable devices (the so called Industry 4.0). The possibility to add sensors to the products, to collect data and to exploit the information underlying has a key role towards innovative devices and services. Data are driving a silent revolution in manufacturing, leading “traditional” industries from the production of merely physical things to a profitable synergy of material and immaterial goods. While data analysis is acknowledged as a driver of the market of services companies, its potential is still to be fully exploited for goods produced by the manufacturing industry. In this context, Moxoff developed machine learning algorithms to identify trends and classifying the states of a system through the analysis of data recorded real-time by electronic sensors. Tecniplast have been developing a new innovative line of cages that supplements the usual devices needed to take care of laboratory animals with movement sensors. The AI algorithms exploit the potential of the data collected by the sensors, to help the final clients of the cages to optimize their core business activities. Facility managers benefit of predictive maintenance tools to enhance the care of the animals. Researchers collect much more information to analyze the animal behavior during their experiments. The presence of cage sensors is a true revolution in the correspondent Laboratory Animal Industry because it leads to provide several new and unique benefits without affecting the current way of housing animals.

Abstract

In this talk I will describe how Pirelli has grown from scratch a Data Science and Analytics department. I will give an overview of the kind of projects, roadmap and Vision we have for Industry 4.0 in Pirelli with a particular focus on the approach we have adopted to create synergies between factories, management and Data Scientists.