Insights from data are essential to many IoT applications. Digitization and exploitation of data in products, services, and processes makes it possible to find important patterns and to perform complex analysis and optimization through computation. Artificial Intelligence approaches are revolutionizing the potential of the depths and value for these insights, as well as the scale at which the benefits can be derived automatically.
The approaches enable new and more powerful, time-series analysis; audio, image, and video analytics; and they open the way for self-learning controls and contextual reasoning to derive actionable insights. In combination with novel speech and gesture based user interfaces, AI can support users in a new, natural way in performing their tasks. The compact form of IoT devices and the added value of AI enable both novel applications and business models. This all makes Artificial Intelligence a quickly evolving area in IoT-related research and in industrial applications.
Within the session we will discuss the potential of novel technologies as well as feasible applications of AI for IoT use cases and the experiences from actual deployments.
The session will investigate topics on:
- Automated Machine Learning
- Machine Learning on the Edge
- Semantic modeling and context learning
- Deep Learning for Control
- Constraint Optimization
- Ethics and Applications
Joern Ploennigs, IBM
Dr.-Ing. Joern Ploennigs is scientific researcher and manager at IBM Research – Ireland. He works on several aspects of enriching IoT by cognitive computing including machine learning, semantic reasoning, and natural interfaces to enable autonomous, highly scalable, and accessible IoT solutions for a sustainable future. Prior to joining IBM, he was heading the junior research group at Technische Universitaet Dresden, Germany, as well as the data analytics group in the Irish strategic research cluster ITOBO as Feodor-Lynen fellow of the Humboldt-Foundation. He holds a master in electrical engineering for automation and control and a PhD in computer science from Technische Universitaet Dresden. He is chair of the IEEE IES TC BACM and board member of the IEEE IoT initiative as well as program committee member of several renowned international conferences and journals.
Barry O’Sullivan, University College Cork
Professor Barry O’Sullivan is an award-winning academic working in the field of artificial intelligence, data analytics, AI and data ethics, constraint programming, and operations research, for more than two decades. He holds the Chair of Constraint Programming in the Department of Computer Science at University College Cork. He served as Head of Department, Computer Science, from 2012-2015. He is the founding Director of the Insight Centre for Data Analytics at UCC, and a Principal Investigator at the Confirm Centre for Smart Manufacturing which is based at the University of Limerick. He is an Adjunct Professor at Monash University, Melbourne, Australia.
Professor O’Sullivan is a Fellow and current Deputy President of the European Artificial Intelligence Association (EurAI), one of the world’s largest AI associations with over 4500 members in over 30 countries. Professor O’Sullivan was President of the International Association for Constraint Programming from 2007-2012. In 2013 he received a UCC Leadership Award and won the Association for Constraint Programming Distinguished Service Award in 2014. He was named Science Foundation Ireland Researcher of the Year for 2016, UCC Researcher of the Year in 2017, and elected to membership of the Royal Irish Academy in 2017, Ireland’s highest academic accolade.
Professor O’Sullivan has been involved in winning over €200 million in research funding, of which approximately €30 million has directly supported his research activities at UCC. He runs his own AI consulting firm, AI Machina, and welcomes opportunities to engage in projects with governments, multi-nationals, public bodies, and SMEs, alike.
Alexander Fay (IEEE Member’02, Senior Member ’07) received the Diploma with honors and the Ph.D. with honors in electrical engineering from the Technical University of Braunschweig, Germany, in 1995 and 1999, respectively. He worked five years at the ABB Corporate Research laboratories in Heidelberg and Ladenburg in the position of Head of Mechatronics research group and Head of Engineering Methods research group, among others. Since 2004, he is Full Professor and Head of the Institute of Automation Technology at the Helmut-Schmidt-University in Hamburg, Germany.
In 2002, he was awarded one of “World’s Top 100 Innovators under 35” by MIT Technology Review. Alexander Fay received the Ring of Honor of the Association of German Engineers (VDI) in 2009.
He is member of the scientific advisory board of the German Society for Measurement and Automation (GMA) and Head of its Engineering and Operation of automated Facilities department. He was a member of the IEEE Industrial Electronics Society Administration Committee 2009-2011. Since 2009, he serves as an Associate Editor of IEEE Transactions on Industrial Informatics.
He is author of more than 100 reviewed national and international journal papers and of more than 150 reviewed national and international conference papers, and of 7 patents. Currently, he has 20 researchers at his institute and is supervisor of about 20 Ph.D. theses. Between 2011 and 2013, he served as the Dean of the Mechanical Engineering Faculty, and between 2013 and 2015 as the Dean of the Industrial Engineering Department of his university.
Since 2014, he is member of the Scientific Advisory Board and of the WG 2 “Research and Innovation” of the German „Industrie 4.0“ initiative..
His main research interests are models and methods for the engineering of large automated systems, especially in the process and manufacturing industries, in buildings and transportation systems. Among other methods, he develops and employs knowledge-based methods, ontologies and other AI techniques. The aim of his team is to develop models, methods and tools to increase engineering efficiency and to assist in the engineering and operation of industrial plants. The research projects are mostly undertaken in close cooperation with industry, i.e. suppliers and/or users of automation technology. Current research topics are, among others
- agent based control of transportation systems (e.g. railway freight transport, baggage transport in airports, and work piece transport within production systems)
- requirement engineering for production systems regarding flexibility, agility, robustness etc.
- development of a reference framework for evaluating and comparing agent-based control algorithms in comparison to other decentralised and central control algorithms under varying degrees of disturbances
- stability analysis of decentralised decision-making entities with time-delayed feedback
- market-based multi-agent system for decentralized power and grid control
- automatic code generation based on GRAFCET specifications
- simulation-based virtual commissioning of production automation systems,
- knowledge-based automation of engineering tasks,
- automated reduction of alarm floods,
- methods for the systematic modernization of automation systems and adaption to “Industry 4.0” needs.
- robot-based manufacturing of lot-size-1-workpieces,
- real-time location and location-based services.
Talk Title: How AI Can Orchestrate IoT Devices in Industry 4.0
The term “Industry 4.0” (or “Industrie 4.0” in German) refers to the fourth industrial revolution, a change in the organisation and management of the entire value chain throughout products’ life cycles due to increased and ubiquitous information and communication technology. Industry 4.0 relies on all the relevant information being made available in real time through networking of all the different components that make up the value chain, as well as the ability to use this data to ascertain the optimal value stream at any given point in time. This networking of human beings, objects and systems enables the emergence of dynamic, real-time optimised and self-organising value chains involving several different companies and providing new business models. These value chains can be optimised based on a variety of criteria such as cost, availability and resource consumption.
The core technologies of Industry 4.0 are IoT on the one hand, as machines, transport systems and even work pieces are equipped with sensors and communication means, and AI on the other hand, as industrial production and transport need to be controlled in a decentral way, where software agents, semantic technologies, and other facets of AI turn out to be helpful.
To structure the field of Industry 4.0, the German Industrie 4.0 Initiative has developed a dozen “Application Scenarios”, covering products, processes and production resources, from engineering to operation. These Application Scenarios help to structure existing and future use cases and application examples. Furthermore, they give guidance towards the need for AI technologies, both regarding functional and non-functional requirements.
Within this talk, the requirements and the successful application of AI technologies in selected Industry 4.0 Application Scenarios will be discussed, putting emphasis on how AI is employed to implement the ideas of IoT in the production and logistics domain.
The author is a member of the Working Group 2 of the German Industrie 4.0 initiative, which has elaborated these Application Scenarios and developed a research roadmap from this endeavor. Currently, the Application Scenarios are matched with similar approaches on an international level, e.g. with Italy, France, and Japan.
Alex J Joseph
Talk Title: AI in Industry 4.0
Industry 4.0 drives collaboration and networking across the factory value chain for the increased ability to connect and manage devices; and for reason of near real-time data collection which leads to a whole lot of insights of what is happening. While this entire concept will enable the Enterprises to become better connected it will mean the traditional architecture to converge and flatten as solution providers port or re-write existing applications to run on top of IoT platforms. IoT enabled MES, sensors, instrumentation, controls, assets, and materials will be the new norm; where new business models for Smarter Factories will be around descriptive, predictive and prescriptive analytics – which will be the difference maker. Smart Factories need to be designed with much resilience in planning, to ensure zero tolerance to disruption. IBM IoT Solutions offer a wide range of AI models to drive cost savings and operational efficiency across the factory value chain. Energy, being limited, and at the same time for being one of the major expense in any Industry, will need to be treasured for optimal (or sustainable) usage. These are another significant perspective of a Smarter Factory, which is driven from the context of Energy Optimization – again driven by the right use of AI models.
Michele Magno is currently a Senior Researcher and lecturer at ETH Zürich, Switzerland at the Department of Information Technology and Electrical Engineering (D-ITET) . He received his master and Ph.D. degrees in electronic engineering from the University of Bologna, Italy, in 2004 and 2010, respectively. He was postdoctoral research at Tyndall Institute, Ireland and University College Cork, Ireland, and a visiting professor at University of Nice and ENSSAT University of Rennes, France. His current research interests include smart sensing, low power machine learning, wireless sensor networks, wearable devices, energy harvesting, low power management techniques, and extension of the lifetime of batteries-operating devices. He has authored more than 140 papers in international journals and conferences. Some of his publications were awarded as best papers in IEEE conferences such as IEEE International Conference on E-health Networking, Application & Services 2018, IEEE Sensors Applications Symposium (SAS) 2018, IEEE International Workshop on Advances in Sensors and Interfaces 2017 among others. He is a senior IEEE member and an ACM member.
Talk Title: Smart Sensing in the IoT Era: Machine Learning on Ultra Low Power Microcontrollers
Machine Learning (ML) and artificial intelligence are pervading the digital society. Today, even low power embedded systems are incorporating ML, becoming increasingly “smart.” This talk gives an overview of ML methods and algorithms to process and extract useful near-sensor information in end-nodes of the “internet-of-things”, using low-power microcontrollers ARMCortex-M; Bluetooth low energy SoC, and other low power microcontrollers. The main objective is to show how Machine Learning algorithms can be adapted to the performance constraints and limited resources of low-power microcontrollers. Finally, the talk will introduce an open source library based on Fast Artificial neural network, that allows easily to implement Artificial Neural Network on ARM Cortex-M Family.
Prof. Robert Shorten graduated from UCD with a B.E. degree in EE in 1990 and a Ph.D. degree in 1996. Since then, he have worked in academia and industry and will continue to split time between both.
In industry, he has been employed by both Daimler-Benz Research and IBM Research, where he led the Control/Optimization activities at IBM Research Ireland in the area of Smart Cities, as well as being a consultant to industry.
Highlights of Robert’s academic activities include time as a Professor at the Hamilton Institute, where he was a co-founder of the Hamilton; spending time as a Visiting Professor at TU Berlin in 2011-12, and a very formative period working with Professor Narendraat Yale University.
Robert is currently a Professor of Control Engineering and Decision Science at UCD and continues to act as a consultant to IBM Research.