Intelligent sensor systems or self-learning machines will revolutionize many industries. The energy industry too is increasingly seeing an opportunity to cut costs and provide another level of service.
Governments, companies and consumers have now understood that more intelligent energy networks are essential for the future energy infrastructure and thus for economic growth. At the moment, some of the electrical networks (grids) in use are over 100 years old. And they are inadequate given our increasing thirst for energy and the decentralized feed-in of alternative energy sources.
At the same time, more and more energy consumers are producers too – so-called prosumers. This makes it even more difficult to ensure that all customers can be supplied with all of the energy that they need at all times. This requires huge investments in smart grids, investments that can barely be refinanced through network fees.
In order to ensure that the operating costs remain manageable, the networks must be made artificially intelligent. Only then will it be possible to make reliable predictions regarding electricity consumption in households and companies. According to a study carried out by Sopra Steria Consulting, it should be possible to reduce the costs of network management and energy trading by around 20 percent by using predictive intelligence solutions.
However, only ten percent of energy suppliers currently use systems for predicting consumption and network capacity use. Half of the companies in the sector would like to address this problem as soon as possible and see great potential as a result of AI developments. At the same time, 60 percent of companies are planning a significant increase in the number of intelligent sensors that they use in order to reduce system maintenance costs.
Predictive maintenance saves money
Further savings though AI technologies can be achieved with predictive maintenance. In the process, intelligent local network stations provide data regarding transformer temperatures, load profiles, energy flows and switching statuses. This is then used as a basis for predictive system maintenance.
According to the Sopra Steria Consulting survey, 60 percent of energy suppliers regard this as the key benefit of artificial intelligence. The development of suitable sensor systems is therefore right at the top of the to-do list. 60 percent of those surveyed are planning a significant increase in the number of intelligent sensors that they use in order to reduce system maintenance costs.
As far as know-how in the field of artificial intelligence is concerned, the energy suppliers consider themselves to be in a good position. Owing to the development of renewable energy sources, the energy sector has been involved with intelligent technologies for developing smart home and smart grid solutions for years. Those within the sector regard the immature nature of some of the technology and the use of artificial intelligence for purposes other than network management as the biggest challenges. Such purposes include replacing routine work with robotic process automation, enhancing customer service with digital assistants or complementing chatbots with artificial intelligence.
IBM Watson for smart grids
In order to better predict supply and demand patterns for electricity in smart grids on the basis of historical and current weather data, ABB and IBM recently entered into a strategic partnership. The digital ABB Ability offering with the cognitive abilities of IBM’s Watson IoT will help to predict how much energy is required and how much can be produced. The energy suppliers will then use the predictions as a basis for load management and real-time pricing.
Of course, the use of artificial intelligence in smart grids is not entirely new. Back in 2011, the “Peer Energy Cloud” project was launched. The project came up with entire activity profiles for individual households in Saarlouis on the basis of data regarding power consumption at individual power sockets.
As a result of this, the load flow could be optimized and a virtual marketplace for electricity trading within a so-called “micro grid” was set up. A consortium of partners made up of the German Research Center for Artificial Intelligence (DFKI), the Karlsruhe Institute of Technology (KIT), AGT Germany, Seeburger AG and Stadtwerke Saarlouis was behind the project. Details of other “smart” micro grids can also be found. However, the great wave is yet to materialize. The current hype regarding artificial intelligence could provide a new boost for smart grids.