Contents
SpecificationASAM

ASAM SOVD

Service-oriented vehicle diagnostics

ASAM SOVD is the specification for a web-style diagnostic API that lets testers reach high-performance computers and other service-oriented vehicle software, while classic control units stay on UDS.

What it is

ASAM SOVD (Service-Oriented Vehicle Diagnostics) is the ASAM specification for diagnostic access to vehicle computers that expose software as services. A client such as an engineering tool, a workshop tester, or a remote backend sends web-style requests over Internet Protocol to an SOVD server. The server reports the entities it hosts, their faults and data, and the operations it allows. The specification targets high-performance computers and zone controllers, and it leaves classic diagnostic services for traditional control units to their own standards.

Who it applies to

SOVD applies when an OEM or supplier puts a diagnostic server on a node that speaks IP and hosts vehicle functions as separate software. Diagnostics owns the capability list and the tester requirements. Software on the high-performance computer or zone controller implements the server. E/E architects decide which nodes use this path. A control unit that only speaks classic bus diagnostics is out of scope unless a specified gateway presents it through SOVD.

What it covers

The specification defines the API and the information a server makes available. It does not define the network schedule or the functional-safety case.

Topic What you use it for
Discovery Find the entities a node exposes, such as applications, components, and groupings, before reading data.
Faults Read and manage fault information those entities hold.
Data Read current values and structured data that would not fit a short classic diagnostic frame.
Operations Invoke only routines the entity allows, such as a self-test or another controlled action.
Bulk data Transfer logs and other large payloads.
Clients and roles Decide which clients may connect, because the API is reachable over IP.

Resource names, status codes, and payload schemas are in the purchased edition. Pin one edition so clients and servers are not mixed across revisions.

How it fits with other documents

ISO 14229 (Unified Diagnostic Services) stays the protocol for classic ECUs. SOVD does not redefine those services, so vehicles commonly use UDS on traditional controllers and SOVD on the high-performance computer.

ISO 13400 (Diagnostic communication over Internet Protocol) carries UDS over IP. SOVD also uses IP, but the payload is the SOVD API, not a UDS message inside DoIP. A DoIP session alone is not an SOVD session.

ASAM ODX (MCD-2 D) is the data description many UDS tools load. An SOVD server advertises its capabilities through the API. Programs that keep both stacks usually keep ODX for UDS controllers and a separate description for SOVD.

AUTOSAR Classic diagnostic modules remain the path to UDS. An SOVD server is typically built on an Adaptive platform or a custom HPC stack, and using AUTOSAR does not by itself make a node SOVD-capable.

Add the API to the vehicle cybersecurity case: which clients may call it, from which network, and which operations can change safety-relevant or security-relevant state.

In practice

Give each node one primary diagnostic path: an SOVD server on computers that host functions as services, and UDS on other ECUs. If one workshop tool must see both, specify a dual-protocol client or a gateway with tested translation rules.

Agree the entity map before software is frozen: which applications and components appear, which faults and data they own, and which operations are allowed outside the plant. Share that list across diagnostics, software, and architecture. Every state-changing operation needs an owner and a precondition.

The API uses IP, so include the switch, the gateway, and any remote backend. Keep the service off networks where a diagnostic client should not appear, and bring up authentication with the server.

Record the edition in the component requirement. v1.2 is in preparation, so stay on the edition in force until clients and servers are updated and retested together.

What it means for your discipline

  • Diagnostics & lifecycleCore

    You own SOVD for service-oriented nodes. Daily work is the entity map, fault and data content, allowed operations, and the workshop tests for every computer that exposes a server.

  • HPC & application softwareCore

    You implement the server and the software entities behind it. Each release must expose the agreed faults, data, and operations and stay on the baselined edition of the API.

  • E/E architectureCore

    You decide which nodes speak SOVD and which stay on classic UDS, including any gateway between them. That split belongs in the E/E architecture, not in a later tool choice.

  • CybersecurityPartial

    The API is an entry point on IP. Include it in the threat analysis, limit which clients and roles may call it, and keep it off networks where diagnostic access is not intended.

  • Networks & communicationPartial

    SOVD traffic crosses Ethernet switches and gateways rather than only a classic diagnostic connector. You need the routing and port plan that lets an authorized tester reach the server.

  • Base softwarePartial

    Classic basic software keeps its UDS stack. On an Adaptive platform or a custom high-performance computer, you host the SOVD server beside platform services and feed it from the applications.

  • Software update & OTAPartial

    Update campaigns are specified in other documents, though a server may expose update-related operations and bulk data. Agree which of those calls a campaign may use.

  • Functional safetyPartial

    Some operations can reset a function or change state a safety concept relies on. Check which calls are reachable in the field and which preconditions the safety case requires.

No direct stake: Hardware & silicon