Diplomatic Protocol for Chip and Code ✛ ✎

Original: "Evaluating Hardware Abstraction Layer Concepts for Software Defined Vehicles: Insights into Applicability and Effectiveness"
arXiv:2607.00039 · 2026-06-29 · arXiv non-exclusive · independent analysis · 3 min
A comparative analysis of hardware abstractions for software-defined vehicles: lessons from IT, industry, and the automotive sector itself.

Stuttgart, 2025. A software engineer plugs a laptop into a test car, not for diagnostics, but to replace the car's 'personality'—like flashing a smartphone's firmware. But instead of a quick update, he faces a cascade of incompatibilities. Each electronic control unit (ECU) speaks its own dialect, tied to a specific supplier's chip. Software written for a Bosch braking system won't understand a Continental processor without a complex 'interpreter'. This interpreter is the hardware abstraction layer (HAL), a diplomatic protocol between the world of silicon and the realm of algorithms. Without it, the car will never become a software-defined vehicle (SDV)—a machine whose functions are defined by code, not soldering.

The QNX Hypervisor spends less than 2% of CPU time on its own overhead, like a diplomat at a summit devoting only a couple of seconds per hour to translation.

Researchers at the University of Stuttgart, like seasoned negotiators, staged a review of HAL solutions. They didn't limit themselves to the automotive realm but peered into neighboring 'states': smartphones with their Android Treble, the industrial automation of Industry 4.0 with its Asset Administration Shell, the network virtualization of NFV, and cloud interfaces like VirtIO. Two diplomatic approaches crystallized: a lightweight messaging protocol (middleware) or full isolation of the 'embassy' inside a virtual machine (hypervisor). Automotive candidates (AUTOSAR Adaptive, Android Automotive, QNX, EB Corbos, and others) were evaluated against 14 criteria (from determinism to Machine Learning support), sorted by the MoSCoW method into must-haves, should-haves, and could-haves. The maximum score is 45, reflecting the protocol’s maturity.

The results sharply delineated the diplomatic ranks. Hypervisor-based solutions (QNX, EB Corbos, RTA-LWHVR) unequivocally lead in safety and readiness for over-the-air (OTA) updates. Their overhead is nearly imperceptible, and the critical loop (steering, instantaneous hazard response) is shielded by hardware armor. The middleware camp falls short on determinism but offers flexibility and portability: applications for video stream analysis or self-learning can easily migrate between chips. A curious outlier (SOAFEE, a standard yet unborn) theoretically scores a perfect 45 points, promising to encompass everything.

The cold numbers revealed an architectural rift familiar to diplomats since the days of John von Neumann: how to combine the reliability of strict separation with the evolutionary flexibility of open interfaces. Today, no single HAL is universal. The solution is a hybrid protocol. The critical loop (braking, steering) lives under hypervisor isolation, while everything else, from infotainment to cloud telemetry, communicates via standardized middleware, much like web services over HTTP. This diplomatic pact ensures that a navigation glitch can't affect the brakes and allows the infotainment system to be updated over the air without rebuilding the entire firmware.

The Stuttgart engineer opens his laptop again. This time the protocol works: the car accepts the new code in seconds, silently reconfiguring its behavior. The diplomatic success of the HAL determines whether millions of cars become full citizens of the Internet of Things or remain closed 'iron islands'. Tomorrow, this hybrid interpreter could head into series production — first on the pages of the Eclipse SDV standard, and then onto the roads.

The QNX Hypervisor spends less than 2% of CPU time on its own overhead, like a diplomat at a summit devoting only a couple of seconds per hour to translation.

In the TV series Knight Rider, the car KITT could receive updates and new features 'over the air'—decades before engineers seriously tackled OTA. Today’s work brings that science fiction closer to everyday reality.

Concepts
support vector machine electric vehicle