AI helps to predict renewable energy generation and manage smart grids that monitor and control electricity flow, manage energy storage and prevent failures & cyber-attacks
The electrical engineering community has been at the forefront of experimenting, using, and developing AI technologies to improve electrical components, equipment, machines, and systems. AI is used to improve their design, performance & optimization, adaptive & fault-tolerant control, fault detection & diagnosis, condition monitoring & predictive maintenance, reliability, and cyber security. As early as the 1960s, expert systems were developed for power system scheduling, machine analysis & design, power electronics systems, etc. In 1990, the IEEE Control Systems Magazine published a special issue on ‘Neural Networks in Control Systems’. According to Google Scholar, over 10.7 lakhs, 6.6 lakhs, and 6.5 lakhs of electrical engineering-related papers, reports, and patents include the words ‘artificial intelligence,’ ‘neural network,’ and ‘machine learning,’ respectively. AI infusion in electrical engineering is growing rapidly. For example, the market for AI in the energy sector is expected to increase five-fold by 2030 and the market for smart lighting and control system is likely to double within the next five years.
AI is used to predict and manage electricity demand and supply, optimise power plant operation, detect faults, and predict power system behaviour. It also helps to predict renewable energy generation and manage smart grids that monitor and control electricity flow, manage energy storage, and prevent failures and cyber-attacks. AI is helping in demand-side management by adjusting appliance usage in response to price or supply changes, reducing peak demand. It can help to optimise building energy consumption by learning occupants’ behaviour and adjusting energy usage accordingly. It can be used to optimise the operation of electric vehicles and charging stations. Chatbots can help power utilities in customer interaction. Natural Language Processing (NLP) can be used to analyse maintenance and injury reports for improving maintenance and operations. Computer vision can automatically detect violations of safety and other work protocols. Power utilities use intelligent robots and drones for inspecting, cleaning, and maintaining high-risk, time-consuming, and difficult-to-access assets. The global market for robotics in the renewable energy sector alone is expected to reach 11.6 billion by 2025. AI along with advanced communication networks, phasor measurement units, and sensors can implement the self-healing capabilities of the smart grid. Development of intelligent digital twins, virtual replicas of physical machines and systems, is increasing for monitoring, analysing and optimising physical assets’ performance and energy consumption, predicting and preventing failures, and improving reliability, efficiency, and safety, while also reducing costs and environmental impacts.
Interdisciplinary education for preparing Electrical AI Engineers
The challenges of Industry 4.0 and Circular and NetZero economies call for interdisciplinary engineering solutions. Industry 4.0 involves integrating physical, biological, and digital worlds to create new technologies and systems, while the goals of Circular and NetZero economies require increased environmental focus through policies like green credits and renewable energy certificates. AI, a worthy daughter of electrical engineering, is being integrated with contemporary computing and electrical technologies to ensure an uninterrupted, high-quality electricity supply, increase energy efficiency, and mitigate negative environmental impacts of electricity generation, distribution, and consumption.
An Electrical AI Engineer combines knowledge of electrical engineering and AI to develop intelligent systems to support the planning, design, manufacturing, operation, and maintenance of electrical components, systems, and infrastructure. Either electrical engineering or ‘computer science and engineering’ graduates can pursue these roles. It requires the electrical engineering curriculum to also include some relevant computing topics like digital instrumentation and control systems, embedded systems, IoT, robotics, M2M communication, computational methods for electric engineering, smart grid, etc., along with generic computing and AI. At the same time, the CSE or AI programs should also include a few topics related to physical engineering disciplines, e.g., electric and mechanical workshop, machines, sensors, instrumentation and control, and also some in related computing, e.g., embedded systems, IoT, robotics, computational engineering analysis, CAD, simulation, etc.
In order to groom engineers for successful careers in the Industry 4.0 era, the engineering curriculum should offer a multidisciplinary education including their core engineering discipline with a balanced exposure to other engineering fields, computing technologies, and non-engineering subjects. The ‘computer science and engineering’ programs too must supplement ‘computer science’ courses with physical engineering-related courses. Further, the creation of active-learning-based interdisciplinary educational engagements is even more crucial. This requires an overall transformation of engineering education and the dismantling of silos between physical engineering and computing and fostering of inter-disciplinary team teaching by multiple departments.