Waymo LLC has secured a major milestone in automotive technology with a newly patented system for comparative vehicle simulations. This innovation focuses on the patent titled Using simulations to identify differences between behaviors of manually-driven and autonomous vehicles. The patent describes a sophisticated method designed to bridge the performance gap between human drivers and self-driving systems.

Advancing Autonomous Reliability

Waymo LLC has been recognized with the Swanson Reed Patent of the Month award for April 2026 within the Automotive, Battery and Self-Driving Technologies sector. This accolade highlights the critical importance of the invention in establishing safety benchmarks and operational parity between human operators and artificial intelligence. By leveraging historical log data to recreate exact real-world scenarios within a virtual environment, the system provides an objective lens through which autonomous software can be refined and validated against human intuition and reaction times.

The methodology involved in this patent is considered outstanding because it addresses the core challenge of quantifying autonomous vehicle progress. Rather than relying on generic simulations, Waymo utilizes specific lane points and path segments traversed by manual drivers to create a direct comparison. This level of granularity allows developers to pinpoint exactly where autonomous logic differs from human behavior, enabling targeted improvements in smoothness, safety, and efficiency. It serves as a vital tool for the industry to prove that self-driving technology is not just functional but superior to traditional driving methods.

Winning this award underscores Waymo leadership in the self-driving industry during a period of rapid technological evolution. As regulatory bodies and consumers demand higher transparency regarding AI safety, this invention provides a verifiable framework for performance assessment. The ability to extract metrics from both live log data and simulated runs ensures that every software update is rigorously tested against real-world human baselines, making it an indispensable asset for the future of transportation and mobility.

Patent Abstract

A simulation may be used to determine a difference between progress of a manually-driven vehicle and progress of a simulated autonomous vehicle. The method includes retrieving log data collected for the manually-driven vehicle driving along a route, generating a plurality of path segments for a portion of the route. The plurality of path segments corresponds to points in a lane that the manually-driven vehicle traveled through on the portion of the route. The method also includes running, using a software of the autonomous vehicle, a simulation of the autonomous vehicle driving along the plurality of path segments, extracting metrics from the log data and the simulation, and determining the difference between a first progress of the manually-driven vehicle and a second progress of the simulated autonomous vehicle based on the metrics.

U.S. R&D Tax Credit Compliance

This invention meets the R&D tax credit rules in the USA through the four-part test which requires the activity to be technological in nature, intended for a permitted purpose, aimed at eliminating technical uncertainty, and involving a process of experimentation. Waymo development of this simulation technology constitutes a systematic effort to improve the functionality and safety of autonomous vehicles through rigorous iterative testing. Each simulation cycle represents a technical experiment aimed at overcoming the uncertainties of how AI handles complex road conditions compared to human drivers. The work relies on principles of computer science and engineering to create a more reliable commercial product.

Practical Applications for R&D Tax Credits

• Application 1: Comparative Metric Framework: The development of scripts and algorithms that automatically extract and normalize comparative metrics from disparate log data sets constitutes a process of experimentation. Engineers must iteratively test various mathematical models to ensure the simulation accurately reflects real-world physics and human behavior patterns.
• Application 2: Path Segment Reconstruction: Creating the plurality of path segments based on manual driving points requires solving technical uncertainties regarding data mapping and spatial accuracy. This work is technological in nature as it relies on complex geometric modeling and sensor data fusion to recreate a high-fidelity virtual route.
• Application 3: Safety Benchmark Optimization: Using simulation software to determine progress differences serves the permitted purpose of improving the reliability and safety of a commercial autonomous driving system. The development of these automated benchmarking tools qualifies as R&D because it aims to achieve a significantly more advanced level of performance over existing manual baselines through systematic trial and error.