Overview of the Concepts and Functionalities of OCTANE

The IOSB team and their partners are developing a freely available simulation platform, which is based upon three essential concepts: A modular and easy-to-expand software architecture, an intuitive and gradual usability, and the use of levels of detail.

The special feature of the OCTANE concept is the individual adaptability of the simulation to particular requirements concerning computational effort, result quality, usability and available data. This is ensured by a modular and expandable software architechture, as well as by an intuitive usability adapted to various user backgrounds, and by the use of independent levels of detail. Moreover, due to the free availability of the source code, OCTANE provides the user with the possibility of making any adaptions of the simulation, integrating any new functionality and examining the underlying simluation models.

An overview of OCTANE can be found in the topic points below.

  • Modular Software Architecture

    The software architecture of OCTANE was designed in a way that extention and adaptions of the simulation are easily possible on the source code level. The flexible configuration of complex scenarios as well as the extension with new simulation models and even the integration of entirely new simulation kinds is designed to be as easy as possible with the architecture.

  • Intuitive Usability

    The usability concept of OCTANE is meant to offer both occasional users and experts a fast way to produce and simulate scenarios. The different concepts of usability are intuitively built upon each other, facilitating a smooth transition from a beginner's mode (graphical user interface (GUI)), over the use of adapted simulation scripts, up to the expert’s mode (using OCTANE as code library). Ultimately, modifications of the open source code itself permit almost arbitrary adjustments of the simulation.

    Examples:

    • GUI
    • Script-based batch simulation
    • Source code programming
  • Usage of Levels of Detail

    In order to optimally adjust computational effort and result quality to the particular application, the simulation models are realized in different levels of detail for enabling an independent scalability of separate submodels.

    Examples:

    • Primitive engine and sophisticated sensor models for testing a sensor fusion algorithm
    • Sophisticated vehicle dynamics with a primitive radar model for testing an ACC control
  • Vehicle Dynamic and Vehicle Modelling

    Vehicle dynamics in OCTANE are simulated with a combination of rigid body physics and automotive models. For this reason a complete simulation model of a vehicle is implemented, composed of the essential vehicle components. It is possible to simulate any number of vehicles of different accuracy in a single simulation.     

    Examples:

    • Engine
    • Tires
    • Suspension
    • Chassis
  • Advanced Driver Assistance Systems

    A full range of common advanced driver assistance systems (ADAS) is already integrated in OCTANE, for instance in order to realistically simulate emergency braking with an anti-lock braking system (ABS).

    Beispiele:

    • Anti-lock braking system (ABS)
    • Adaptive cruise control (ACC)
  • Sensors

    The different sensors are modelled at different levels of detail. Depending on the kind of sensor and its level of detail, dynamic maps of the surroundings are directly retrieved from the simulation (in the sense of a perfect sensor), or raw data for algorithmic processing is either simulated by the GPU or by raytracers for high accuracy.

    Examples:

    • Camera
    • LiDAR
    • Radar
    • Ultrasound
    • Car2X communication
  • Environment

    The modeling of the environment incorporates a multitude of possible factors influencing the simulated vehicles, for example road surfaces with different friction parameters, or weather conditions like crosswind, fog or rain. Traffic flow behavior, for instance modeling the degree of capacity of road segments and thereby influences the individual vehicle, is also included in this category.

    Examples:

    • Weather conditions
    • Time of the day
    • Road surface models
    • Traffic flow
  • High Level and Low Level Driver Behavior

    OCTANE provides the possibility of simulating abstract traffic flows on road networks, individual vehicles by standard models, as well as realistic drivers given their environmental perception. A simple driver model for controlling traffic participants following a navigation goal is currently under development.

  • OpenStreetMap Data Import und Manipulation

    In addition to manually generating a road network, it is possible to import OpenStreetMap data (OSM data) for creating road networks as well as 3D-environments including buildings. A generation of terrain and vegetation in combination with OSM data for simulating different terrain profiles is currently being pursued.

    Examples:

    • Generation of streets and sidewalks from OSM data
    • Random terrain generation for OSM data
  • Visualisierung

    Visualization

    The visualization includes an intuitive graphical user interface (GUI), a live-view of the simulation and sensor data, and the possibility for displaying different simulation parameters. For a more complex visualization for advertising purposes for instance, interfaces to professional renderer engines will be available.

    Examples:

    • Live view
    • Plotting functionality
  • Interfaces

    Connecting other software or hardware, for example external maneuver planning algorithms or USB steering wheels will be possible through interfaces.

    Examples:

    • Interface to USB steering wheel and pedal
    • Interface to professional renderer engines
    • Matlab data export
  • Platform Independence

    OCTANE is currently developed platform independently and shall be released for Windows and Linux, in addition to being server-compatible.