Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Internet of Things (IoT) is an emerging domain that promises ubiquitous connection to the Internet, turning common objects into connected devices. The IoT paradigm is changing the way people interact with things around them. It paves the way for creating pervasively connected infrastructures to support innovative services and promises better flexibility and efficiency. Such advantages are attractive not only for consumer applications, but also for the industrial domain. Over the last few years, we have been witnessing the IoT paradigm making its way into the industry marketplace with purposely designed solutions. In this paper, we clarify the concepts of IoT, Industrial IoT, and Industry 4.0. We highlight the opportunities brought in by this paradigm shift as well as the challenges for its realization. In particular, we focus on the challenges associated with the need of energy efficiency, real-time performance, coexistence, interoperability, and security and privacy. We also provide a systematic overview of the state-of-the-art research efforts and potential research directions to solve Industrial IoT challenges.

Industrial Internet of Things: Challenges, Opportunities, and Directions

The Internet of Things (IoT) is entering the daily operation of many industries; applications include but are not limited to smart cities, smart grids, smart homes, physical security, e-health, asset management, and logistics. For example, the concept of smart cities is emerging in multiple continents, where enhanced street lighting controls, infrastructure monitoring, public safety and surveillance, physical security, gunshot detection, meter reading, and transportation analysis and optimization systems are being deployed on a city-wide scale. A related and cost-effective user-level IoT application is the support of IoT-enabled smart buildings. Commercial space has substantial requirements in terms of comfort, usability, security, and energy management. IoT-based systems can support these requirements in an organic manner. In particular, power over Ethernet, as part of an IoT-based solution, offers disruptive opportunities in revolutionizing the in-building connectivity of a large swath of devices. However, a number of deployment-limiting issues currently impact the scope of IoT utilization, including lack of comprehensive end-to-end standards, fragmented cybersecurity solutions, and a relative dearth of fully-developed vertical applications. This paper reviews some of the technical opportunities offered and the technical challenges faced by the IoT in the smart building arena.

IoT Considerations, Requirements, and Architectures for Smart Buildings—Energy Optimization and Next-Generation Building Management Systems

The smart metering and communication methods used in smart grid are being extensively studied owing to widespread applications of smart grid. Although the monitoring and control processes are widely used in industrial systems, the energy management requirements at both service supplier and consumer side for individuals promoted the evolution of smart grid. In this paper, it is aimed to disclose in a clear and clean way that what smart grid is and what kind of communication methods are used. All components of a smart grid are introduced in a logical way to facilitate the understanding, and communication methods are presented regarding to their improvements, advantages, and lacking feature. The developing generation, transmission, distribution and customer appliances are surveyed in terms of smart grid integration. The communication technologies are introduced as wireline and wireless classification where the key features are also tabulated. The security requirements of hardware and software in a smart grid are presented according to their cyber and physical structures.

A survey on smart metering and smart grid communication

Wearable salivary uric acid mouthguard biosensor with integrated wireless electronics

Smart grid implementation requires that multiple distributed systems are synchronized within and between substations. This is needed to coordinate the Intelligent Electronic Devices (IEDs) as well as to align in time the measurement data associated with the same event, but collected in different points of a broad geographical area. The time synchronization accuracy requirements in power systems are application-specific and are classified in the IEC Standard 61850-5:2013. Depending on such requirements, alternative solutions can be used. For instance, sub-microsecond time synchronization is essential for the implementation of Wide Area Monitoring, Protection, and Control (WAMPAC) systems. In fact, time synchronization directly and indirectly affects Phasor Measurement Units (PMUs) accuracy. The PMUs can measure the magnitude, the phase, the fundamental frequency and the Rate of Change of Frequency (ROCOF) of ac voltage and current waveforms at times synchronized to the Universal Coordinated Time (UTC). Among such quantities, the phase data are those most strongly affected by time synchronization. If time synchronization accuracy is within ±1 µs, its impact on synchrophasor phase measurement is in the order of a fraction of mrad, i.e., low enough to have a negligible impact on the state estimation uncertainty of most of transmission and distribution systems. However, the need to interpolate the missing phasor values when data from PMUs with a different reporting rate are collected and aligned in time may unexpectedly boost synchrophasor phase estimation uncertainty.

Time Synchronization for Smart Grids Applications: Requirements and Uncertainty Issues

A simple, portable, and low-cost system for noninvasive measurement of vital parameters, based on two sensors conveniently combined in a unique compact-in-size and low-power structure, has been presented and characterized in this work. The proposed solution allows both sensors to fit in a small structure, in order to be applied on the earlobe (or analogue body part of a person) in a clip-like fashion, making it suitable for portable and battery-operated systems. The two involved sensors are as follows: a photoplethysmographic (PPG) sensor, used to optically measure the change in volume of blood in the arteries, thus to acquire information about heart beating, and a sensor composed by two electrodes, used to perform an impedance measurement, thus to monitor the dielectric properties of biological tissues and fluids, like blood flow.

Characterization of Multi-sensor System for Noninvasive Measurement of Vital Parameters

The standard IEC 61850 describes the communication inside Substation Automation System (SAS) setting the base for new interoperating devices that share information using Ethernet network infrastructure. More recently, the request of time synchronization among the SAS components for timestamping of critical events or synchrophasor measurement, generated a very high interest in the IEC 61850 working group for the distributed synchronization protocols like IEEE 1588. This paper analyzes the current situation, matching different implementations of the IEEE 1588, and their respective accuracy, with the requirements of typical SAS applications: from the milliseconds synchronization accuracy required by fault recorder application to the µs required by the Sampled Values. The main focus of the paper is on the integration of IEEE 1588 with real already-existing systems and the goal is the performance comparison of different architectures. For this reasons, experimental tests have been carried out using a real SAS network infrastructure (redundant gigabit optical fiber Ethernet network, managed using Rapid Spanning Tree Protocol, RSTP ). The results show that, under ideal conditions, the IEEE 1588 protocol can achieve synchronization accuracy slightly worse (80ns) than legacy IRIG-B systems (15 ns). However, IEEE 1588 makes possible a great cabling reduction and a simplified maintenance when synchronization is transferred directly on the SAS System Bus. On the other hand, experiments highlight that nodes connected through a legacy RSTP infrastructure, which does not support IEEE 1588 messages, can lose the time synchronization in the case of network traffic or reconfiguration of the network. Therefore, the integration of IEEE 1588 over existing IEC 61850 Station Bus is basically a trade off between installation costs and performance, since it requires the replacement of already installed Ethernet infrastructure. In the last part, the paper analyses the challenges for the next SAS generation that has to combine on the same plants the high accuracy IEEE 1588 implementation with high-availability protocols based on IEC 62439. Since, the lifetime of a SAS is more than ten years and the adoption of these technologies takes several years, some solutions that introduce synchronization and high availability protocols guaranteeing the compatibility with existing technologies, have been analyzed, too. For instance, an adaptation block that, (when attached to a high-availability network) reconstructs the interface signals (e.g. synchronization signals) required by traditional Bay, has been proposed.

Behavior and Future Challenges

Retrieving system logs is one of the main sources of information about system operation and a widespread solution adopted in computer science for control management and maintenance purposes. In this paper an automatic log analysis technique, based on the training of a neural network classifier, has been applied to two different scenarios in the smart grid and industrial Internet of Things (IIoT). The methodology is described and validated using real data from a northern Italy utility company and a SIMATIC S71500 PLC installation. Results show how the approach can work in both scenarios with thousands of devices for smart grids and up to 50 devices sending up to 50 messages per seconds in the IIoT scenario with reasonable low bandwidth consumption to preserve the timing characteristics of the industrial real time ethernet.

Applying Automatic System Log Analysis to Industrial Automation Systems for IoT Integration

Low Power Wide Area Networks (LPWANs) are wireless technologies well accepted in the Internet of Things (IoT) scenarios because they operate in both public and private infrastructure using unlicensed frequency bands. Differently from other wireless standards, the LPWAN target applications are low-rate, sporadic-transmission, low-energy systems. Surely, fast response time and high reliable communication are out of the original LPWAN scope. In this work, a first investigation about the use LoRaWAN (one the most diffused LPWAN protocols) for notification of emergency messages across industrial plants is reported. The opportunity of using recently proposed low-cost repeaters, and the creation of a clever replication pattern of radio messages are the keys of the proposed approach. The results of the preliminary analytical feasibility assessment are presented, confirming the possibility to increase success ratio of notification and respective confirmation in case of emergencies.

Stefano Rinaldi

Papers

Time Synchronization for Smart Grids Applications: Requirements and Uncertainty Issues

Characterization of Multi-sensor System for Noninvasive Measurement of Vital Parameters

Behavior and Future Challenges

Applying Automatic System Log Analysis to Industrial Automation Systems for IoT Integration

Investigating redundancy of LoRaWAN for emergency notifications in industrial plants