reviewing AI origins and multi-domain applications
1950: The magazine Mind publishes the article “Computing Machinery and Intelligence,” written by Alan Turing. The British mathematician presents a test to analyze a machine’s ability to exhibit intelligent behavior similar to or indistinguishable from that of humans. Currently, artificial intelligence is being applied in multiple sectors such as health, finance, transportation, education, art, and energy. One of the artificial intelligence techniques that have had a major impact is deep learning, specifically artificial neural networks, inspired by the human brain and designed to simulate how neurons behave when exposed to sensory input such as sound or images.
AI for power Systems: BD4OPEM H2020
2020-2023: In the European project Big Data for Open Innovation Energy Marketplace (BD4OPEM H2020), coordinated by UPC, applications of artificial intelligence are being developed for electrical grids. Through machine learning and optimization techniques, value is extracted from operational and non-operational data for enhancing the electrical system decision-making pro- cesses, especially at the distribution level. Electrical distribution grids are evolving towards smart grids thanks to an increasing integration of renewable sources, electric vehicles, and energy storage systems. All these elements are facilitating the decarbonization of the system but also introduce complexity in the operation and management of the networks.
In BD4OPEM, data-driven tools are enabling the development of services such as topology analysis, observability, measurement error detection, generation and demand prediction, flexibility prediction in different time horizons, congestion detection, fraud detection, and distribution network planning, among others. The output information from these services, either individually or combined, is im- proving the operation, maintenance, and planning of the electrical grids. These services are accessible from BD4OPEM Marketplace (access here) and are being tested in 5 pilot sites: Estabanell Energia (Spain), OEDAS (Turkey), Elektro Celje (Slovenia), Brussels Health Campus (Belgium) and Nuvve (Denmark). The project is now reaching its final stage and this newsletter is focusing on the main achievements and lessons learned.
Looking to the AI future after BD4OPEM H2020
Considering the technological advancements in computational capacity, big data, cloud computing, the Internet of Things, and the progress of artificial intelligence in its multiple applications, it can be deduced that it will have a transformative role in society. However, it can never replace humans. Our responsibility goes beyond the ”simple” creation of algorithms based on artificial intelligence; it extends to their use and supervision.
The magazine ”Business Insider” published an article in 2020 titled ”The Role of Humans in the Era of Artificial Intelligence: This is What the Great Leaders Think.” Chinese entrepreneur Jack Ma reflected as follows: ”We are smarter; we were the ones who invented computers (…) I have not seen a computer invent a human being yet.” And he stated, ”Computers only have chips; humans have hearts.” It is in our hands to make rational use of the potential of artificial intelligence. It depends on us to live in Aldous Huxley’s ”Brave New World.”
bd4opem marketplace video
We are glad that this newsletter has three demo videos to demonstrate the functionality of the BD4OPEM Marketplace!
Demo Video 1:
Data contracting: Access to the BD4OPEM Marketplace to search and contract a new data set. The demo video shows an example scenario of a user, namely a Data user, accessing the marketplace to search for a new data set to contract, and the process of contracting it from the point of view of the Marketplace, the Data user, and the Data provider.
Demo Video 2:
Service contracting: Access to the BD4OPEM Marketplace to search and contract a new service.
The demo video shows an example scenario of a user, namely a Service user, accessing the marketplace to search for a new service to contract, using also the data set contracted previously, and the process of contracting the service from the point of view of the marketplace, the Service user, and the Service provider.
Demo Video 3: A service user is accessing the Marketplace to launch the contracted service.
The demo video provides an example of a Service user using the service and the outcome obtained from the service.
learnings from work package 4
LESSONS LEARNED REGARDING DATA INFRASTRUCTURE:
Having a cloud-based infrastructure has contributed to the project’s flexibility and scalability, allowing Data Providers to adapt quickly to changing project requirements and efficiently manage resources.
LESSONS LEARNED REGARDING DATA PREPROCESSING:
It is essential to prioritize data harmonization and standardization. This allows for seamless data integration and sharing, improving collaboration between Data Providers and Energy Services, and allowing that one service can be tested in more than one demonstration site easily.
Investing resources in data cleansing and pre-processing techniques improves data quality and, consequently, Energy Services performance. This involves identifying and resolving inconsistencies, duplicated records, outliers, and missing data points that could affect the accuracy and performance of services.
Through big data processing and analysis tools, the services can use and analyze large volumes of data in an adequate processing time. This enables some energy services to make real-time data-driven decisions.
LESSONS LEARNED REGARDING DATA PRIVACY AND SECURITY:
Implement robust data privacy measures, including encryption and anonymization techniques, protects sensitive information during data collection, storage, and Energy Services usage. Compliance with relevant data protection regulations, such as the General Data Protection Regulation (GDPR), is essential.
Establish robust access controls and user authentication mechanisms for the BD4OPEM Marketplace, ensuring only authorized associates can access and manipulate sensitive data and Energy Services of the project.
MORE LESSONS LEARNED:
It is highly beneficial for the optimal development of services to have diverse actual data sources from different countries, as it helps improve their performance. Furthermore, by developing services with actual data from the pilot sites (Data Providers), the gap for future implementation in the actual market is reduced, as the results obtained have been validated in a similar environment to the real one, using actual data.
Ensuring proper documentation through the deliverables regarding the energy services description and methodologies helps to maintain transparency and reproducibility. This facilitates knowledge transfer within the BD4OPEM-involved partners.
work package 8
We effectively used our social media channels and engaged our audiences with interview videos, shared news from the BD4OPEM partners, pictures, collaborations with other BRIDGE-related projects, and shared project results.
The regular releases of pilot site interview videos have generated a lot of interest and engagement on social media, showing that big data is also something that is relevant for people in the end.
It is with great pleasure that we now will collaborate with one of our sister projects, SYNERGY, with planning and holding the final event which will be on the 14th of November in Barcelona. More information will come.
Smart City Expo World Congress 2022 – SCEWC
SCEWC 2022 was hosted in Barcelona, Spain from the 15th of November – the 17th of November. BD4OPEM attended the event, hosting a booth and interacting with other exhibitors and visitors interested in the Energy system of the future.
Enlit 2022 took place in Frankfurt, Germany between the 29th of November – the 1st of December. BD4OPEM attended the event, hosting a booth in the EU Project Zone. The project was presented to other EU-funded projects and interested visitors of the exhibition. Monica Aragués Penelba presented the project on stage. Several partners attended the exhibition in order to learn more about the current commercial and technological developments within the energy system sector.
European Big Data Value Forum 2022 – BDVA
BDVA was hosted in Prague, Czech Republic on the 21st November – 23rd November. The project attended the event by hosting a booth and presenting on stage. (ATHOS?) presented results of the project.
Daniel J Brandt Dennis Lindbom
pilot site - Anell
PLC (Power Line Communication) is one of the technologies that has allowed data communication to greatly evolve. This technology is mainly used to establish the connection channel between Data Concentrator Units (DCU) and Smart Meters to retrieve data. However, certain types of noises can disrupt the communication channel, especially as non-linear loads increase, such as photovoltaic plate converters and electric vehicle chargers. Although background, narrowband, and impulsive noise have been mitigated in some extent, there is a type of impulsive noise that has not yet been mitigated, to the point that it is the cause of communication drops.
As a solution to these communication problems resulting from the Spanish pilot of BD4OPEM, the startup Energy in the Cloud, developed a filter behind the meter which can cancel this impulsive noise.
This filter, called NOCA-BP (Fig-1), allows PLC communication used by smart meters to report consumption to data concentrators for billing, and it is installed in parallel to the smart meters in an easy-to-install solution compared to the traditional way of installing in series with the low voltage line and the loads. By connecting the NOCA-BP filter in parallel with the network, an attenuation of said interference is achieved, restoring the communication of the meters.
This filter is valid for the new PLC PRIME v1.4 standard (PRIME Alliance Technical Group, 2014) which allows extension in the FCC (10-490 kHz) or ARIB (10- 450 kHz) bands. Specifically, it makes use of the frequency band between 42 and 472 kHz divided into 8 channels that can be used as independent channels or several of them together as a single transmission/reception band. Since the study of the behavior of this new standard against noise in these higher bands has not yet been carried out, the analysis that will be carried out in this project acquires relevant interest.
Lluís Cànaves Navarro Jordi Jene
Figure 1 NOCA-BP-Installed Filter in Aněll grid
Work package 9
Through the development of D9.2 together with WEP and the partners we have explored two possible revenue models for the key exploitable results. Pay per-use and yearly subscriptions. And we conclude that the pay-per-use model is too volatile and difficult to estimate to be a viable model, and as such, that the traditional subscription-based model is to be preferred.
The life cycle assessment is progressing with Dominik Huber from VUB leading the task. We can see a path where the BD4OPEM LCA can adapt and use new types of methodologies to assess the positive societal impact that for e.g. V2G can have, the work is still ongoing but there are exciting insights to come.
WP9 together with WP8 has performed a vast number of stakeholder engagements and we can see a common trend among energy grid stakeholders; digitalization is de facto the future and there are a lot of solutions being developed within that realm (flexibility, V2G, topology, etc.) – and they all face the same challenge with having access to and working with ”clean” big data. It is here the BD4OPEM marketplace can provide the best value and connect energy grid stakeholders.
Daniel J Brandt