IoT and AI – Stimulants for power system automation


Abhijit Maitra, Armatrics Applications



Today’s buzzword in the industry is Internet of Things (IoT). That extended to industries are also feeling the heat in the name of Industrial IoT (IIoT) and so called Industry 4.0. The digitization of the power sector cannot escape the disruptive wave that is coming with Industrial IoT. The transformation is directed to impact the energy value chain. The IoT ecosystem as a product of the Digital Transformation renders technological innovations in the fields of Connected devices, Augmented reality, Cloud computing, Bigdata and Decentralized ledger.  With the plethora of data emanating from these IoT enterprises, comes the need for analysis through Artificial Intelligence (AI) techniques and extending the technologies to carry out varying degrees of autonomous action by Robotics. The combination of artificial intelligence and automation termed Intelligent automation — is already helping achieve unprecedented levels of efficiency and quality for companies. Advances in machine learning techniques, improvements in sensors, new generation of hardware and software robots and ever-greater computing power have potential excitements to deliver in various ways in the power industry and destined to transform operations of strictly regulated, conservatively managed power entities and systems at the generation, distribution as well as the consumers and utilization fronts.


1. What is IoT and IIoT?


The concept of Internet of Things evolved some time back. Coinage of the phrase-term ‘Internet of Things’ (IoT) dates back in 1999 to describe this vision (now reality) of Internet-connected sensors, devices and citizens. Concept of smart technologies and universal connectivity in diverse fields are emerging as evidenced in Smart cities, Smart environment, Smart water, Smart metering, Smart surveillance, Smart logistics, Smart agriculture, Smart buildings, Smart manufacturing…  etc. What the Internet of Things (IoT) is targeting is to bridge the cyber and the physical worlds and all this is going to challenge conventional practices – the way we live and do business.


Consequently, the business organizations which include the utility sector like Power will readjust the conventions to mitigate the waves of changes deploying the IoT system. This disruptive change, as perceived by industry pundits, is taking realistic shape and gaining maturity at a fast pace.


The Internet of Things has been largely bifurcated between consumer andindustrial in terms of adoption, investment, and ecosystem since its inception. While consumer IoT brings multitudes of convenience in lifestyle with its applications, Industrial IoT (IIoT) addresses efficient and friendly manufacturing or services. The fallout of consumer IoT is inconvenience, whereas for IIoT it can lead to critical and grave consequences.


In industry we have long experienced machine to machine connectivity, and the M2M concept is nothing new. IoThas resemblance to M2M in that it connects devices to networks. But M2M have historically been associated with specific and problem-drivensolutions, primarily in closed systems. The IoT instead refers to a shift of perspective in that it extends the internet to provide as a means of communication between things that could benefit from the connection. How? Not merely due to ongoing technology development, but it aims at a paradigm shift in problem solving and innovation approach whereby the IoT frequently utilize the existing communication infrastructure and feeds from open information, mobile networks and cloud solutions, creating many new areas to explore.


Industry 4.0, as the Germans prefer to call this digital transformation actually emphasizes 4th generation of Industrial revolution, where the current trend of automation and data exchange in manufacturing technologies includes cyber-physical systems, the Internet of things and cloud computing.


2. IoT, AI – the next leap in industrial automation


We are more or less all familiar with the conventional industrial automation systems, which comprisesat its heart Supervisory control and data acquisition (SCADA) systems, Human machine interface (HMI) , Programmable Logic Controllers (PLCs), Distributed Control systems (DCS), other Controllers and Gateways, interconnecting network on Ethernet TCP/IP, Fieldbus, Profibus etc. interfacing with sensors, instruments, actuators, drivers and other output devices. The result is an intelligent manufacturing infrastructure for increased safety and efficiency and lowered costs.


At its core, the present day automation aims to bring together the advances of two transformative revolutions: conventional industrial automation, as cited above, that arose and graduated from the Industrial Revolution; and 


the powerful advances in computing, information and communication systems established via the Internet Revolution. Theutilities and a host of energy companies are increasingly interacting with new technologies such as IoT, Artificial Intelligence (AI), blockchain and robotics.

IoT deals with real-time optimization of production and supply chain networks in an industrial setting by networking sensors, actuators, control systems and machinery together. It also automates the process controls, service information systems and operator tools using digital controllers inorder to achieve enhanced productivity and safe distribution system. To establish the same, it leverages the communication network available through Internet to access wide based information and dataas a key manipulating driver. 

The combination of artificial intelligence (AI) and automation termed Intelligent Automation — is already helping achieve unmatched levels of efficiency and quality for companies. Artificial intelligence and automation are hardly new, but the technologies have matured substantially in recent years.Intelligent automation systems try to streamline decision making. The complex information and the automated analysis and decision process,which typically use tools forcollecting, collating, extracting and analyzing information,are embedded in a workflow after human review, validation and approving machinedecisions.An increasing number of tools are being created that allow machines to become more sophisticated in how they learn and make decisions.Machine learning refers to the ability of computer systems to improve their performance by exposure to data without the need to follow explicitly programmed instructions. At its core, machine learning is the process of automatically discovering patterns or prototypes in data. Once discovered, the pattern can be used to make predictors.

The other emerging area of business process automation is distributed ledger technology through blockchain.Blockchain is a distributed database that securely maintains a growing ledger of transactions (defined as transfers of data natively tracked by the system) and other information among participants in the network. Developers are attracted to blockchain for its inherent security, data integrity, decentralized nature, and its ability to simultaneously provide both public openness and effective anonymity. Bitcoin was the first implementation of blockchain, which proved its robustness and overall value.

3. Growth and Complexities before the power sector

While it was possible to run a generating plant with a handful of gauges in earlier days, it is quite impossible to run large turbo-generators today without automatic control system and automation are common using specialized computers for stress evaluators, permissible load calculators, automatic run-up controls, logging and graphic systems through power SCADA. Several levels of controls involving complex arrays of devices are used to meet the requirements of the power system. The transmission controls consist of power and voltage control devices, such as static VAR compensators, synchronous condensers, switched capacitors and reactors, tap-changing transformers, phase-shifting transformers and HVDC transmission controls. The prime objective and challenge of satisfactory operation of the power system against this complex scenario is to stabilizesystem voltages and frequency and other system variables within their acceptable limits through Generation side control and Demand side control against varying conditions of system load and losses and maintain power interchange with adjoining control areas.

For a number of reasons however the complexity of power systems is increasing and new problems are arising for human supervisions and manual control operations. Moreover, the power companies are increasingly burdened to supply abundant energy for the growing needs of power, throwing challenges not only in technology but also in management, finance and manpower.

To meet all these challenges, smart system infrastructure, smart generation, and smart utilization are essential. Achieving this will require a fundamental shift from the traditional unidirectional flow of energy and communication to a bidirectional power flow where the receiving end potentially can return power flow to the grid based on consumption level and its own generation capacity. Power supply systems of today and tomorrow must integrate every type of power generation to bridge the increasing distances between power generation and the consumer.A distributed utility configuration is believed to be the present and best approach.

However, with the addition of more renewable generation to the grid, there are also unintended consequences and inherent operating challenges.Because of the small scale and dispersed configuration of its components, a distributed utility approach to system planning potentially may encounter more resistance from the public than would to build a few remotely sited plants.Widespread adoption of distributed resources could face opposition from an ingrained utility culture.

In a nutshell,

  • The legacy Grid is increasingly inadequate and insecure and in the midst of growing competitions
  • Distributed generation, storage and management throwing more changes and complexities
  • Pressures on costs and revenue, finance and manpower
  • Social context of use of electricity against a commercial approachof utility companies often resultin hazardous or non-environment friendly usage of labor, fuel, land or geography.

4. Approach for improvements in power sector

It is well understood that there is high cost and high impact on society as a result of transmission disturbances and blackouts and vulnerability of power networks to terrorist attacks. Transmission Systems Operators (TSOs) are already focusing on organizational structures, procedures, and technical innovations that could improve the flexibility and robustness of power systems and achieve the overall goal of providing secure power supply. The industrial IoT, AI and associated toolsemploying vastly improved Information and communication technology are trending fast to the roles of immediate solution hubs.

Distributed Utility approach

Traditional approach uses aggregated load forecasts and the installation of large central generation plants, whereas a distributed approach is interested in local conditions and costs.  This entails a more fine-grained approach to system planning, and therefore a distributed utility approach has to access and process a greater amount of information related to the area and time specific costs of the power system. The utility agency has to make a rigorous analysis of the system’s area and time specific costs to evaluate risk potentials and eventually derive result from the analysis about how and when to invest for capacity upgrades, so that resources are planned for right size and location in the overall power system.  The distributed utility approach aims to capture a host of potential benefits, including a reduced reliance on the grid, reduced environmental consequences, lower costs and improved reliability when compared with traditional power generators.

Digital transformation

In industrial context, increasingly all areas of operation and management are getting shaped by the onslaught of digitization.

  • Substituting technology in operations (OT) to manage the complexities of technological demands and monitor events, processes and devices and make adjustments in enterprise and industrial operations to enhance quality of output.
  • Leveraging information technology (IT) based ondata-centric computations in management, finance and manpower for decision making and increasing efficiency

Such efforts employing “IT-OT convergence” place strictly regulated, conservatively run power sector entities in the midst of fast-moving ICT(Information and communication technology) trends, such as cloud computing, Big Data, the Internet of Things and Artificial Intelligence leading to transcend conventional performance tradeoffs into overcoming challenges and achieving unprecedented outcomes.

Legacy applications for grid operations are generally not equipped to handle even the increase in data from today’s smart meters and sensors, leaving even larger scope of data usage for optimal grid performance. When compared with the big data being collected by utilities on the distribution side, the potential data volume from power plants is some orders larger in magnitude.As increasing numbers of sensors can gathermore and more plant data to integrate with data in historians, business office systems, and external data feeds (weather, market prices, and so on), the need for seamless integration of those data streams for faster operational and business decisions become very important. By now, cloud computing is a well-known tool for a wide variety of industries as is also establishing its footprints in the power sector, capable of offering solutionin –

  • Renewables management –
    • Short term renewable forecasting
    • Improving equipment maintenance.
    • Atmospheric observations of wind, solar and estimation for storage energy
  • Demand management –
    • Demand response algorithms
    • Enhanced feedback of energy consumption & performance,
    • Learn and anticipate user behavior to optimize consumption.
  • Infrastructure management –
    • Digital asset management with machine learning algorithms
    • Highlight risks and opportunities across infrastructure
    • Model possible scenarios, advise on action and impacts.

Smart Grid

The major area where IoT deals with energy management systems is the Smart Grid. The Smart Grid features using IoT technologies include Advanced Metering Infrastructure (AMI), SCADA (Supervisory Control and Data Acquisition), Smart Inverters, Remote control operation of energy consuming devices. Instead of overloads, bottlenecks and blackouts, Smart Grids will ensure the reliability, sustainability and efficiency of power supplies. As a result, automation will increase significantly; information and communication systems within the network will be systematically extended and harmonized. Smart substations will reduce dependence of manpower with respect to planning and operation requirements and comprehensive monitoring on a continuous basis will improve the way that plants and the grid are run.

Smart Metering

Smart Metering systems integrate a number of innovations: digital technology, communications, control and better operation of networks. Smart Metering technologies change the way that metering works completely. They provide customers with much more information on how they use energy and enable those customers to reduce their usage.  Customers are benefited by information on energy costs and carbon emissions, energy.Multi tariff functions with smart meters allow demand response techniques and allow electrical appliances to be automatically controlled. Smart metering helps utilities to improve prediction, monitor and track renewable power and streamline power consumption, reduce operating expenses by replacing manual operations, Improve customer service through profiling and segmentation and reduce energy theft.

Robotics and AI

Drones and Robotic inspections enable operators to remotely access hazardous and inaccessible zones in a safe and cost-effective manner, especially in nuclear power plants. These technological developments have positive impact on O&M in the power industry. Machine learning, which is an important part of artificial intelligence, has huge potential e.g. many supervised machine learning models can be used for time series forecasting. Regression models can directly forecast electricity generation, consumption and price. Classification models can forecast the probability of a spike in electricity prices.

Distributed ledger and blockchain

Blockchain technology shows a lot of promise. Other than being used to execute energy supply transactions, it could also provide the basis for metering, billing and clearing processes. Other possible areas of application are in the documentation of ownership, the state of assets (asset management), smart contracts, and guarantees of origin, emission allowances and renewable energy certificates.

5. Use cases

The benefits that are visible due to this paradigm shift in technology as a result of the digital transformation drive is slated to impact the power sector dramatically over the next decades. Advancements in AI, distributed ledgers and robotics will impact utilities as these trends combined with the changes in the energy sector functioning like  distributed energy resources, increased proliferation of sensors on infrastructure and behind the meter devices and demand management improvements will unleash a variety of transformative use cases in the sector. For example, devices which auto-detect demand levels on the grid and reduce power flow could be driven by AI and registered by blockchain.

Very few plants have moved to cloud-based or any sort of real predictive maintenance, as still most plants do reactive maintenance. This area has a huge potential as is exemplified by the remote Monitoring and Maintenance being done by Siemens for more than 9,000 turbines (wind and fossil-fueled) online. Every day, the gas turbines generate some 26 GB of data while wind turbines generate 200 GB. Remotelymonitoring turbines provide multiple benefits, including longer service intervals and predictive maintenance and maintenance planning, which can lead to increased profitability forcustomers. For wind turbines, Siemens says it can provide remote remedies for 85% of all alarm situations.

Utility companies in Australia, Estonia and Sweden has tied up with Ericsson and have started building basic Smart Grid communications and smart metering infrastructures, and are now looking at further ways to maximize possible uses for the new data.To get the benefit of such applications as grid monitoring and control, metering, asset management and tracking and field force communication, the IoT connected devices in energy sector plays a big role. These devices range from meters, grid sensors and actuators to energy boxes and electrical appliances. Smart meters monitor energy consumption at regular intervals and transmit information to the utility companies.  Apart from the benefit of helping make a resilient grid system,  consumers are also benefitted having the provision of accurate, up-to date billing information, which allows the utilities to introduce more flexible payment methods. Smart meters also eliminate the need for periodic trips to each physical location to read a meter.Meters in the form of in-home displays (IHDs) allow consumers to monitor and better control their usage of electricity. Home area networking (HAN) communication technologies such as ZigBee, allows smart meters to function as “home connected meters” supporting home automation connectivity with lighting, appliances and security systems. Now, further outreach to reap benefits from extensive use of smart meters are in the pipeline like improved detection of grid fault location, identification and restoration, local energy loop optimization and creating more efficient market mechanisms – both wholesale and local.

Very recently, digital industry solutions specialist GE has launched digital power plant systems for gas and coal plants. In new plants, GE’s technologies have increased the average conversion efficiency from 33 percent to 49 percent.For the longer-standing coal plants, efficiency improvements are substantially less, although emissions of greenhouse gases can be reduced by 3 percent. These efficiency gains come about through a clever blend of Internet of Things (IoT) technology and active monitoring. Optimizing fuel combustion, tuning the plant to adjust to the properties of the coal being burned, adjusting the oxygen levels in the boiler, and reducing downtime due to equipment failures all have an impact.

Google has put its DeepMind artificial intelligence unit in charge and using AI to manage power usage in parts of its huge data centers. The result of this experiment is a 40 percent reduction in the amount of electricity needed for cooling, which Google describes as a “phenomenal step forward.”

In New York state, neighbors are testing their ability to sell solar energy to one another using blockchain technology. In Austria, the country’s largest utility conglomerate, Wien Energie, is taking part in a blockchain trial focused on energy trading with two other utilities. Meanwhile in Germany, the power company Innogy is running a pilot to see if blockchain technology can authenticate and manage the billing process for autonomous electric-vehicle charging stations.

6. Challenges of the technology

With the new technology, new set of challenges are evolving. Discovering the appropriate technology for usage to improve operations, innovate and grow is a strategic decision fraught with challenges. Another critical challenge is staffing and training plans to develop talent pool to handle changed skill requirements, job descriptions and organizational models. With development of a more complex automation system through IoT and AI, it’s basically an invitation to new sets of risks. Assessing risks and deploying mitigation measures on cyber threats, privacy breaches, product liability are big challenges by itself.

7. Conclusion

Irrespective of the industry, the basic trend that is touching across the board is the new technological ammunitions in the form of internet of things, artificial intelligence and its allied tools in the strive for advancing automation to the next level. The fraternity is gung-ho on the whole endeavor, even if the flipside offers potential pitfalls. The physical replacement of human intervention by using process automation along with simple decision making roles has been in existence for some time now. The automation which addresses more convolutedand multivariate decision making aspect uses these new technologies which aim to surpass the human limitations to its maximum, while every caution is being taken to mitigate the challenges on road to implementation of the same. The power sector as it is one of the most complex system, undoubtedly has every reason to benefit in a big way from these trends in the industry. The decision support algorithms from the humongous amount of data generated from such implementation is slated to churn out effective results, and will be a necessary aspect of the future automation industry and for power sector in particular.

Leave a Reply

Close Menu