Deep Learning (DL) algorithms are an extremely promising instrument in artificial intelligence, achieving very high performance in numerous recognition, identification, and classification tasks. To foster their pervasive adoption in a vast scope of new applications and markets, a step forward is needed towards the implementation of the on-line classification task (called inference) on low-power embedded systems, enabling a shift to the edge computing paradigm. Nevertheless, when DL is moved at the edge, severe performance requirements must coexist with tight constraints in terms of power/energy consumption, posing the need for parallel and energy-efficient heterogeneous computing platforms. Unfortunately, programming for this kind of architectures requires advanced skills and significant effort, also considering that DL algorithms are designed to improve precision, without considering the limitations of the device that will execute the inference. Thus, the deployment of DL algorithms on heterogeneous architectures is often unaffordable for SMEs and midcaps without adequate support from software development tools. The main goal of ALOHA is to facilitate implementation of DL on heterogeneous low-energy computing platforms. To this aim, the project will develop a software development tool flow, automating: • algorithm design and analysis; • porting of the inference tasks to heterogeneous embedded architectures, with optimized mapping and scheduling; • implementation of middleware and primitives controlling the target platform, to optimize power and energy savings. During the development of the ALOHA tool flow, several main features will be addressed, such as architecture-awareness (the features of the embedded architecture will be considered starting from the algorithm design), adaptivity, security, productivity, and extensibility. ALOHA will be assessed over three different use-cases, involving surveillance, smart industry automation, and medical application domains

To unleash the full potential of IoT, realizing the digital society and flourishing innovations in application domains such as eHealth, smart city, intelligent transport systems, and smart manufacturing, it is critical to facilitate the creation and operation of trustworthy Smart IoT Systems. Since smart IoT systems typically operate in a changing and often unpredictable environment, the ability of these systems to continuously evolve and adapt to their new environment is decisive to ensure and increase their trustworthiness, quality and user experience. The DevOps movement advocates a set of software engineering best practices and tools, to ensure Quality of Service whilst continuously evolving complex systems and foster agility, rapid innovation cycles, and ease of use . Therefore, DevOps has been widely adopted in the software industry . However, there is no complete DevOps support for trustworthy smart IoT systems today. The main technical goal of ENACT is to develop novel IoT platform enablers to: i) Enable DevOps in the realm of trustworthy smart IoT systems, and enrich it with novel concepts for end-to-end security and privacy, resilience and robustness strengthening trustworthiness, taking into account the challenges related to “collaborative” actuation and actuation conflicts. ii) Facilitate the smooth integration of these to leverage DevOps for existing and new IoT platforms and approaches (e.g., FIWARE, SOFIA, and TelluCloud). This will be accomplished by evolving current DevOps methods and techniques to support the agile development and operation of smart IoT systems, and provide a set of novel mechanisms to ensure quality assurance and trustworthiness, such as actuation conflict handling, continuous testing and delivery across IoT, edge and cloud spaces and end to end security and privacy management. Through this ENACT will provide a DevOps framework for smart IoT Systems

The most challenging applications in heterogeneous cloud ecosystems are those that are able to maximise the benefits of the combination of the cloud resources in use: multi-cloud applications. They have to deal with the security of the individual components as well as with the overall application security including the communications and the data flow between the components. The main objective of MUSA is to support the security-intelligent lifecycle management of distributed applications over heterogeneous cloud resources, through a security framework that includes: security-by-design mechanisms to allow application self-protection at runtime, and methods and tools for the integrated security assurance in both the engineering and operation of multi-cloud applications. The MUSA framework leverages security-by-design, agile and DevOps approaches in multi-cloud applications, and enables the security-aware development and operation of multi-cloud applications. The framework will be composed of a) an IDE for creating the multi-cloud application taking into account its security requirements together with functional and business requirements, b) a set of security mechanisms embedded in the multi-cloud application components for self-protection, c) an automated deployment environment that, based on an intelligent decision support system, will allow for the dynamic distribution of the components according to security needs, and d) a security assurance platform in form of a SaaS that will support multi-cloud application runtime security control and transparency to increase user trust. The project will demonstrate and evaluate the economic viability and practical usability of the MUSA framework in highly relevant industrial applications representative of multi-cloud application development potential in Europe. The project duration will be 36 months, with an overall budget of 3,574,190 euros.