D.3.3: IT environment Specification and Architecture

The aim of the ICT4CART project is to design, implement and test in real-life conditions a versatile ICT infrastructure that will enable the transition towards higher levels of automation. It focuses on four high-value use cases: 1. UC1 - Smart parking and IoT services, 2. UC2 - Dynamic adaptation of vehicle automation level based on infrastructure information, 3. UC3 - Intersection crossing (urban) and lane merging (highway), 4. UC4 - Cross-border interoperability between Austria and Italy. The above use cases will be implemented and tested in four test sites in Austria, Germany, Italy, and at the Austria-Italy border. This document is the first version of the IT environment specification and architecture (task T3.3 of ICT4CART). It defines the scope of the ICT4CART IT environment and provides an initial specification of the functions and components thereof. The document provides a high-level architecture as well as more detailed architectures focused on the various components. The ICT4CART IT environment architecture covers data exchange and management aspects, data analytics, semantic interoperability, environment perception models (EPM), and high-precision positioning. In order to design a generic architecture, we started by identifying the commonalities between the various use cases in terms of the data flows. This further allowed us to identify the common architectural components and data exchange standards to comply with across the test sites. The functional viewpoint of the IT environment architecture foresees two types of functions: Core IT Services: These are low-level functions that are common across use cases and pilot sites. As such, they require a high level of standardisation. Core services include cooperative intelligent transport system (C-ITS) services, EPM building and fusion, service provider gateways, geo service, location correction, message brokering, and analytics. Applications: These are services that are specific to the use cases. They typically make use of the core IT services. Applications include, but are not limited to, automation level road clearance, smart parking, and fleet management. The IT environment architecture should cope with various connectivity requirements (e.g., latency, throughput, etc.). Therefore, it should be distributed over both cloud and multi-access edge computing (MEC) infrastructures. While cloud services will provide large-scale and non-time critical services (e.g., service provider gateways, semantic interoperability, geo service, etc.), MEC servers will host localised services requiring low-latency, such as C-ITS services, and EPM services. Such MEC servers are located at the edge of the network, connected to the base stations of the cellular network. They provide computation closer to the vehicles, thereby reducing latency for mission-critical data. As a vehicle is moving, it should remain connected to the closest MEC server to minimise latency. Such vehicle-MEC handover will be facilitated through the use of a geo service that includes a registry of MEC and cloud services with their spatial locations and extents. Environment perception models will be used in ICT4CART to communicate in real time the set of objects involved in the traffic at a given location (e.g., intersection, toll station, etc.). The ETSI TS 103 324 standard specification of collective perception services should be adopted for the encoding and exchange of EPMs. This standard is currently under development. But a number of ICT4CART project partners are involved in this effort. It is recommended in this document that we should consider the current draft of the standard and contribute to its development through the implementation and testing of the ICT4CART use cases