VSSo: Vehicle Signal Specification Ontology Primer

The challenges around the existing variance of describing and working with vehicle data has motivated contributors from the automotive industry to develop and maintain the so-called Vehicle Signal Specification. As the semantics of the specification are limited to a tree-like hierarchy and data types, it has been considered for the foundation of a more expressive model resulting into the Vehicle Signal Specification Ontology (VSSo). This document shall give an overview of the use cases and structure of the ontology.


The goal of this ontology is to provide a reusable model for describing and interacting with vehicle data. It relies on the existing Vehicle Signal Specification (VSS), lifting it to more expressive semantics. As depicted below, the ontology serves as a domain ontology for observations, streams or service interactions.

Vehicle Signal Specification (VSS)

VSS started in 2016 as a project of the GENIVI Alliance aiming at standardizing vehicle signals and attributes. The specification consists of a data model defined in YAML. As shown in the figure below, it resembles a tree structure - not in a sense of a taxonomy, but rather categorization with a part of relationship from the vehicle as a root node to the signals in the leaves. The structure and content is actively worked on and evolves over time. VSS also includes tools that offer serializations in multiple formats (e.g. CSV or JSON).

VSS tree
Basic tree structure of VSS

As VSS comes with a proprietary rule set for the concept definition, VSS falls short, when it comes to connecting signals with other domains.

Use Cases

Goal of VSSo is to be easily accessible and applicable in three different use cases that are clustered into Analytics and Services. The former enables queries about the vehicle's current data or data streams, whereas the latter enables the interaction between other Things and vehicle data.

VSS use cases
Envisioned Use Cases for VSSo

Current Vehicle Data

The core of the ontology is the description of vehicle attributes and signals. From the modeling perspective, we refer to an attribute as a StaticVehicleProperty, that has information about a particular characteristic of the vehicle (e.g. fuel type, brand, etc.). In contrast, we refer to a signal as a DynamicVehicleProperty that is continuously changing over time (e.g. speed, the status of a switch, etc.). Querying such properties' values from a vehicle or finding similarities over a vehicle's fleet is a prominent use case. Let us consider the situation in which the customer wants to remotely see the current status of the vehicle (e.g. door lock, window status, battery status, etc.). If the customer has more than one vehicle in his account, all should be shown. Alternatively, an alarm should be triggered if the doors are locked but the window is open. In all of those cases, a specific value is queried for one or a group of vehicles.

Dynamic Vehicle Data over Time

Another use case is the analysis of dynamic properties over time. Working with the data's streaming-nature is important in situations where dynamic properties' behavior is essential for further analysis (e.g. traffic prediction). Expressing the meaning of data streams requires concepts in the model that refers to sequences and not only isolated observations. It is not realized with the core of the ontology, but instead, it should contribute the domain knowledge to other ontologies (e.g. SOSA, IoTStream). Aggregating dynamic data with high-level concepts help to perform data integration at the edge and facilitates cross-applications interactions.

Interaction with Vehicle Data

The Web of Things Working Group in the W3C develops various protocol and security schemes to describe a device's accessible capabilities and data access. In this sense a vehicle can be seen as a semantic web thing. In this case, the proposed ontology would add the domain knowledge to a well-defined standard.


The structure of the ontology reflects the rules and standard catalogue used and defined in VSS. The rules define the supported kind of nodes (e.g. attributes, signals) and the format of their metadata (e.g. units, datatypes). The standard catalogue uses those rules to define the semantics of the vehicle signals and attributes. A set of tools is then used to serialize the standard catalogue into the desired format (e.g. json, csv, proto, etc.).

VSSo Structure
Structure and components of VSSo

As the nature of the rule set and overall component structure in terms of update frequency and reusability is quite different from the standard catalogue, the ontology is split into:

VSSo Core

The core ontology introduces concepts for the structural elements of VSS defined through the rule set in the specifications. Figure 1 gives an overview of those. The root node is the Vehicle itself. From there the structure is given through so-called Branches. They serve as sorting element for the leaf nodes and are not specified in the specification itself in greater details. The leaf nodes contain the semantic information of signals, which expext to describe and hold information, which changes in greater frequency and of attributes, which are more static. The core of the ontology defines this structure in an OWL ontology and serves as a basis for the defined signals of the standard catalogue and potential further development of the branches as more than structural information.


As the core ontology defined the structure, VSSo holds the vocabulary as defined by the standard catalogue. The main objective is, that VSSo doesn't diverege from the standard catalogue, so this is done automatically through tooling provided in the corresponding repository. The tooling takes the standard catalogue and maps it to concepts defined in the core ontology. The result is an OWL complient ontology, following the standard catalogue of VSS.


VSSo Core and VSSo shall serve mainly as domain ontologies for data exchange and analysis. During the development of the ontologies one objective was to stay conceptionally close to SSN/SOSA.