uah p 18021

Docket: UAH-P-18021

Technology

Integration of advanced driver assistance and autonomous driving capabilities has increased the quantity and complexity of automotive software, making the verification of automotive software updates more challenging. Automakers have few options available for verification of software update correctness prior to deployment. The utility of vehicular and driving simulators in software verification is limited by the fidelity of the simulations to real world hardware and driving conditions. Verification via physical vehicles on a test track is limited by the cost of developing and applying the test scenarios.

Researchers at UAH have developed a technique that leverages multicore and virtualization technologies to safely perform in-situ hardware-in-the-loop (HWIL) verification of a proposed software update by executing the update in parallel with the software currently controlling the vehicle. This approach allows for a side-by-side on-target comparison of code execution to detect anticipated and unanticipated discrepancies in behavior. When an unanticipated discrepancy is detected, a discrepancy report may be forwarded to the automaker for further analysis. 

A key advantage of the proposed approach is that multicore and hypervisor virtualization techniques isolate the execution of the over-the-air software update release candidate, preventing it from interfering with safe operation of the vehicle. This will allow an automaker to integrate customer-owned vehicles into the fleet of physical test vehicles, exposing the proposed software update to the widest possible range of operational scenarios, environmental conditions, and platform variations. This in-situ HWIL verification technique is applicable to other cyber-physical systems including aircraft and industrial control systems.

The technique proposed is readily applicable to other cyber physical systems including aircraft and industrial control systems

Applications

  • In-situ verification of software for a variety of cyber-physical systems including autonomous cars, aircraft, and industrial control systems

Advantages

  • Improved reliability and safety of cyber-physical systems
  • Reduced cost of software verification in cyber-physical systems
  • Ubiquitous HWIL verification across a wide range of real-world operating conditions
  • Ability to control type and volume of feedback to manufacturer
  • Ability to safely incorporate deployed vehicles into fleet of test vehicles
  • Ability to stage verification and deployment of updates
  • Ability to verify performance of
    third-party components

Status

  • State of Development: Proof of concept
  • Licensing Status: Available for licensing
  • Patent Status: Patent pending