Contents

Middleware & runtime on the HPC safety island (MCU)

The safety island uses configured software interfaces and communication paths to exchange control data predictably within the HPC and with other vehicle nodes.

3 min read · Discipline stake: BSW Safety Software Architecture

What runs here

The safety island is the HPC’s real-time MCU subsystem. Its AUTOSAR Classic Runtime Environment (RTE) connects safety application software components; AUTOSAR COM handles configured signal communication toward vehicle networks. Although grouped in this grid cell, the RTE and COM occupy different positions in the Classic architecture: the RTE sits between applications and basic software (BSW), while COM belongs to BSW communication services. See the AUTOSAR Classic Platform architecture.

Component Role
Generated RTE Provides application port interfaces, exchanges local data, and connects runnable entities—the executable parts of components—to configured execution events.
AUTOSAR COM Packs and unpacks signals and signal groups in interaction-layer protocol data units (I-PDUs); applies configured transmission and reception handling.
PDU Router (PduR), adjacent below COM Routes PDUs between communication modules and lower communication interfaces.
Interprocessor communication adapter, where needed Connects the island to application-processor software through a platform-specific mailbox, shared-memory channel, or other transport.

The RTE supports sender–receiver data exchange and client–server operations. Local component communication need not pass through COM. COM reception deadline monitoring must be configured; it is not automatically active for every signal. These behaviours are defined in the RTE specification and COM specification.

Why it looks like this

Safety control needs bounded execution and an understandable path from input reception to output actuation. Generated interfaces, explicit buffers, configured communication periods, and known runnable-to-task mappings make timing analysis and integration review more tractable. Static configuration supports predictability; scheduling, interrupt load, contention, and blocking still determine whether deadlines are met.

The island typically has tighter local memory and processing budgets than the application processor. Selecting only necessary communication features helps control resource use and verification effort. Reusing established Classic components can reduce integration work, but generator configuration and safety evidence remain project responsibilities.

For functions allocated ASIL D requirements, middleware behaviour must support the safety argument. Lockstep execution does not establish software correctness, data freshness, or freedom from interference with the rest of the SoC. Software development evidence must address the applicable requirements of ISO 26262-6:2018.

Safety-island software often changes on a more controlled cadence than application-processor features. Stable interface contracts let feature software evolve while limiting changes to the safety baseline; compatibility still needs verification after updates.

How it connects

Above, control functions and safety monitors access typed RTE ports. Below, execution depends on the configured AUTOSAR OS or supported safety RTOS integration; network traffic continues through BSW interfaces and drivers toward hardware. A typical CAN signal path is application → RTE → COM → PduR → CAN Interface (CanIf) → CAN Driver.

Communication with the application processor requires an explicit integration contract. Shared memory does not by itself define message ownership, atomicity, cache visibility, freshness, or restart behaviour. The transport may bypass COM; document the actual path rather than assuming every island-to-processor exchange is an AUTOSAR signal.

Safety-relevant exchanges need specified integrity checks, timing supervision, and receiver reactions. AUTOSAR E2E protection is integrated at chosen endpoints; ordinary COM handling alone does not provide the complete safety mechanism. Protection boundaries and fault reactions belong in the communication safety design, as described in AUTOSAR’s functional safety measures overview.

Compared with other node classes

  • HPC application processor: Adaptive/SOA middleware supports service-oriented applications; the island prioritizes tightly controlled execution and communication contracts.
  • Zone controller: RTE and COM may accompany extensive gateway routing. The safety island’s scope usually centres on its allocated safety functions and supervision duties.
  • MCU-based ECU: The Classic structure is similar, but an integrated island must additionally account for dependencies on shared SoC resources and processor restart sequences.
  • Smart node / LIN slave: A small protocol handler may provide sufficient communication without a full generated RTE and COM stack.

Best practices

  • Specify units, ranges, validity, maximum data age, startup values, and timeout reactions for each safety-relevant interface.
  • Budget latency across runnable scheduling, buffering, communication processing, transport, and receiver execution.
  • Define shared-memory ownership and coherency rules, including behaviour during independent processor resets.
  • Review generated configuration and maintain compatible application, RTE, BSW, and generator versions.
  • Verify corrupted, missing, repeated, delayed, and stale messages, plus overload and restart scenarios, against required fault reactions.

Common pitfalls

  • Treating lockstep cores or an AUTOSAR implementation as sufficient evidence that the complete communication path meets ASIL D requirements.
  • Reading a retained signal value as valid after its producer has stopped or restarted.
  • Assuming physical core separation removes interference through shared interconnects, memory, clocks, or power.

What it means for your discipline

  • Base software Core

    Configures and integrates the generated RTE, COM, PduR, and processor communication adapters. Daily work includes tracing signal paths, reviewing buffers and callbacks, and resolving configuration compatibility.

  • Functional safety Core

    Defines communication safety requirements, protection boundaries, and fault reactions. Reviews timing evidence and tests that communication failures trigger the required response.

  • HPC & application software Partial

    Implements application components against RTE contracts. Must handle invalid, stale, and unavailable inputs explicitly and respect runnable execution constraints.

  • E/E architecture Partial

    Allocates interfaces between the safety island, application processor, and external nodes. Defines transport ownership, shared-resource assumptions, and restart dependencies.

No direct stake: Cybersecurity · Hardware & silicon · Networks & communication · Diagnostics & lifecycle · Software update & OTA

Where it sits in the software stack

Generic layered software of a central HPC (application processor with hypervisor, plus a safety island) and of a classic MCU-based ECU (AUTOSAR Classic). Highlighted modules matter for this topic. Click any module to open its page.

HPC · application processor (MPU) and safety island (MCU)
  • Classic RTE & COM — Connects safety components through generated RTE interfaces and routes configured signals through COM and PduR.
MCU-based ECU · AUTOSAR Classic

Not directly involved on this node type.