Artificial intelligence (AI) is the driver of the advancing digital transformation for many areas of European society, industry, mobility (autonomous systems), agriculture and also healthcare. In many cases, AI is also one of the key enablers of European industrial competitiveness. However, AI is also essential in other areas, such as in the context of the Green Deal, to realize new smart solutions as well as to farther guarantee our lifestyle and values as a society.
Due to the drastic increase in the amount of data in all sectors, but especially in industry, AI is a basic requirement to continue driving digital transformation. This will support the effectiveness of digitalization and automation technologies in different industrial sectors, while also increasing the resilience of European industry.
The EdgeAI project focuses on the development of AI-based solutions and architectures and is thus aimed at creating sustainable European digital sovereignty. Moreover, the overarching vision and the 6 objectives of the European project Edge AI are exactly in line with the scope of this specific RIA KDT call 2021:
- Develop AI-based edge platforms for end-to-end hardware/software solutions that address the AI design stack and middlerware.
- Provide scalable energy-efficient AI-based edge techniques, methodologies, and frameworks to support multiple OSs and hardware platforms
- Advance multi-core AI-based SoC and SoM microcontroller designs with hybrid architectures, embedded systems, and IIoT devices for industrial environments
- Integration of scalable and modular AI co-design: algorithms, HW, SW, and topologies into the new open architecture HW/SW AI platforms
- Implement reconfigurable AI-based architectures to increase reusability, upgradability/upgradability, and lifetime of AI
- Providing trustworthy and explainable edge AI by design solutions with real-time operational capabilities and dynamic online learning
The Austrian EdgeAI consortium consists of 4 partners. Recognized large companies (IFAT and ams OSRAM AG) are collaborating with TU Wien (Institute of Computer Engineering) and a SME (SoftwareCube) in EdgeAI. The research and development activities in the project, are proven in Austria by 2 demonstrators in the field of Digital Industry. In the project itself, there are a total of 20 demonstrators in 5 value chains, which are precisely aligned with the requirements of SRIA2022. The technology development in EdgeAI as an RIA project ends at TRL5. Therefore, the goal of the Austrian consortium is to validate the two developed demonstrators in their use case environment.
The two use cases are related to the semiconductor industry: ams OSRAM AG is developing methods for Virtual Metrology, which enable real-time process control (a digital twin of the process step) and also show future productivity potentials. In this context, ams OSRAM AG plans to collaborate directly with SoftwareCube, as this company owns important data acquisition and data preparation capabilities of production machines. In collaboration with SoftwareCube and the Vienna University of Technology, IFAT will further develop the knowledge database started in the AI4DI funding project. Natural language processing and causal networks will be used for the digital representation of logical dependencies. The integration of real-time data of the production machines will be done together with SoftwareCube. Additionally, parametric dependencies will be quantified by digitized DoE diagrams together with TU Wien.
Both Austrian research aspects enable meta-learning concepts and the application of predictive data mining to combine the predictions of multiple models. This is especially necessary and relevant when very different types of models are used, as in these use cases.
The highly capital-intensive semiconductor industry has been focusing on automation and digitization for several years and can be seen as a pioneer in these areas. The “Chipsact” which came into force in February 2022 underlines and recognizes the importance of the semiconductor industry in Europe. However, the solutions developed in EdgeAI can also be applied in other industrial sectors, thus recognizing and securing a broader European dimension in terms of sustainability and resilience of production and supply chains in and for Europe.
In addition, EdgeAI is a transnational collaboration of industrial and academic partners within the electronic component and system landscape, including SMEs, to pool AI expertise to achieve outstanding research results. Such close networking and implementation of joint R&D activities is only possible within the proposed KDT initiative.
