In-Tractor Cloud: A vision of Service-Oriented System Design enabled by High-Speed In-Vehicle Networks for a Safer Task and Machine Management

Automotive industry compartment is undertaking a massive technology revolution. ADAS systems and infotainment promise to change the way that customers mean travel and transportation radically, through several use cases. The key enabling technologies for this trend are Ethernet and its newly standardized physical layer, IEEE 802.3bw 100BASE-T1 (a.k.a. BroadR-Reach). From an architectural point of view, the evolution of the applications that rely on Automotive Ethernet resembles in many ways the evolution that the IT has had in the last decades. In the IT world, increased throughput and computational power to the end-user enabled technologies like multimedia streaming; scalability and availability requirements, together with the increased complexity of IT infrastructure, led to the "Anything as a Service" paradigm and Software Defined Networks. For what concerns commercial vehicles, new use cases and the enhancement of some existing ones, require the adoption of an high-speed network like Ethernet. The paper proposes a novel approach that enables new perspectives and applications in commercial vehicles and especially in Agricultural Machines, borrowing the best practices and the background of IT technologies and is intended to be a seminal work to delineate a new concept in the commercial, Agricultural and Heavy-Duty vehicles system design, especially for what concerns the network and services. The work will first review the established IT semantics and architectures, then the possibilities and challenges of the and commercial vehicles applications will be examined, pointing to real use-cases. The emerging paradigm is here called "In-Tractor Cloud". This paradigm brings advantages like scalability, component reuse, HW/SW decoupling and better resource utilization, providing a future-proof architecture. The paper describes a promising system approach where computation, platform, redundancy, are conceived as services, and it shows how this paradigm will enable a very efficient and service oriented management of Security, Privacy, Dependability, Performance and Safety. As an example, the architecture enables implements management through "Major Voting" or simple redundancy, performed using computational power of specialized Tractor ECUs : Implements once connected could ask for remote control function execution as a service provided by the In Tractor Cloud services, in order to increase the confidence on Implement Automation function that involve Tractor power and, in general, its functions. This paradigm is suitable for new commercial vehicles engineering control system designs, and is analyzed with an all-round approach that takes into account as many constraints and requirements as possible.