A total of 12 key exploitable results generated by the project were identified and they have been uploaded to the Horizon Results Platform :
- KonnektBox: A novel fog computing platform enhanced for elasticity and compute continuum paradigms: KonnektBox implements edge level interfaces in charge of deploying the computation across the compute continuum based on the distribution provided by COMPSs. It incorporates a data router service to feed the data analytics methods with data from the geographically distributed sensor in the Florence demonstrator and a distributed storage system (dataClay) to guarantee that the data is available across the compute continuum.
- Vision for the deployment of C-ITS solutions in Florence: Florence gained insight into both mature technologies (as V2X) and cutting edge software architectures which are expected to play a key role in C ITS solutions, supporting the implementation of cooperative, connected and automated mobility (CCAM). Hence, the vision has been updated to encompass C-ITS as well. ELASTIC’s co-design approach has permitted the development of a mature strategic vision for mobility in a complex metropolitan area encompassing functional requirements to challenge the ELASTIC architecture, thus providing inputs for further developments and enhancements.
- COMPSs: has been improved to fully support containerization for a flexible operation in a distributed edge/fog/cloud environment. COMPSs has been extended to support elasticity through the implementation of new scheduling algorithms that adapt the scheduling of analytics tasks, as well as the number of deployed COMPSs workers and utilized computing units.
- dataClay: dataClay has been optimized to share data across different instances and extended to support integration with MQTT and Kafka, improving interoperability with other solutions commonly used in the cloud continuum. dataClay has been integrated with COMPSs NuvlaBox and Konnektbox for the collection of telemetry and the data analytics workflows.
- Software framework for data-analytics workflows: The ELASTIC software architecture is the framework that glues together all the software innovations of ELASTIC, providing the features needed to implement advanced mobility applications. The software framework is based on the seamless integration of the different software components of the ELASTIC software architecture, mainly including COMPSs, dataClay and an advanced distributed monitoring tool that provides real-time information on time, energy, communication and security non-functional properties. The software framework has been validated with a smart transportation application, integrating multiple data sources (cameras, sensors, etc.) to detect several hazardous situations in a natural urban environment, generating real-time alerts to connected vehicles.
- Predictive maintenance for track defects: A maintenance vehicle was equipped with sensors for acquiring the rail profile. The developed software allows to analyse the acquired data to plan maintenance interventions, foreseeing and estimating the wear that could occur in the future in the tramway lines.
- NuvlaBox: NuvlaBox is an edge software that turns any computer into a smart edge device. This is of value in many sectors, including retail, manufacturing, and telecom, where data has to be collected and processed away from the data centre. The NuvlaBox was re-designed to be architecture-compliant with ELASTIC. Additional data managers and tooling were added. This is the unique software on the market which enables devices this way and can then be monitored and updated from a central service which also includes payment and legal terms.
- NGAP: Next Generation Autonomous Positioning (NGAP) demonstrator implements a system for measuring the tram position using only onboard sensors. The system was implemented, installed and tested on trams in an operating environment. NGAP system using only on-board equipment (sensors and edge electronics) can determine the tram position with relevant accuracy. The NGAP system allows the autonomous positioning of the system along the line by using several sensors.
- ADAS (Advanced Driving Assistant System): ADAS implements an on-board system to detect and classify tram obstacles, and it uses onboard radar, LIDAR and camera sensors. It transforms the physical sensors into a smart version through specific methods (e.g. Deep Neural networks (DNN) for video). ELASTIC has permitted to integrate ADAS with: homography to get raw data in a common reference system; Data association link sensors’ measurements to tracks; Data fusion algorithm, based on the multi-target tracking (MTT) approach, tracks detected objects and performs collision checking; the activation of visual and sound warnings to alert the driver is also implemented to complete the solution.
- Safe and Secure Edge Computing Box: The Safe and Secure Edge Computing Box is composed of mechanisms needed to provide a safe and secure execution environment and communication for critical computing nodes. It offers computational resources to applications with the high reliability of mission critical systems and provides the flexibility of services from Cloud Computing. The ELASTIC project permitted the specification of the computing and communication mechanisms needed to host critical applications.
- ICE Knowledge Discovery: This is a software tool for data exploration and visualisation without any predefined model. This tool was born with the project, and was used to help fulfil one of the main use cases in the project.
- Predator + maintenance prediction tool: This new developed tool reads track profiles and indicates to the user where they would need to inspect the trail track due to a detected or predicted worn track. The tool helps rail engineers spot locations where they will focus their attention on maintenance activities. It contributed to the fulfilment of use of the main use cases in the project.