ADAS_Big_Data_Management.jpg

Reducing the number of test drives

ADAS/AD Big Data Management and Analytics Platform

Fully exploit the value of your test drive and simulation data

During the development, verification, and validation of complex automated driving functions, a huge amount of data is generated. This data comes from road testing, vehicle and component testbeds, and from simulations. The measurement data gathered over several projects – and sometimes several years – constitutes a valuable asset. This is especially true if you can analyze this enormous amount of data interactively and with a short response time.
 

Big Data Management Solutions

AVL provides an open and seamless solution for Advanced Driver Assistance Systems (ADAS) and Automated Driving (AD) applications. Key features include the automated and highly scalable execution of analytic scripts on large sets of measurements, including time series and object data. With this feature you can immediately apply new key performance indicators (KPI) on entire historical data.

The output of the automated analytics is fed into a meta database. Here the results (such as detected events) enrich the original metadata of the test drive. This allows the user to interactively query the system. These queries can be expressed highly intuitively, and multi-data source queries are a standard feature of the platform.

Additionally, the user can navigate to events of interest with a simple click and get a variety of synchronized views. These can show videos of the test drive from on-board camera, the time series captured on the vehicle bus, and object data from different perspectives. To represent the environment model, the platform uses the object-oriented description language “Open Simulation Interface” (OSI). This reduces complexity and increases the platform’s compatibility with third party sensors.
​​​​​​​

Key Benefits
 

  • Gain insights from one single source of truth (SSOT) for all ADAS testing data
  • Reduce the number of required test drives by finding relevant events in your existing data set
  • Keep your benchmarks up-to-date by applying newly defined KPIs on your entire historical data set
  • Improve your productivity by reducing the search time for the relevant data
  • Convenient programming interface for ADAS/AD engineers – no Big Data experience required
  • The platform enables users to write simple sequential code in the Python programming language