Here’s how Waymo is using hyper-local weather data to improve the capabilities of its self-driving vehicles –

Author: Eric Walz

Rain, fog, and condensation can affect the performance of an autonomous vehicle’s sensor suite.

Traveling in bad weather can be difficult for some drivers. Driving in the rain, for example, can reduce visibility, fog up vehicle windows, or cause unwanted reflections on wet road surfaces.

But for self-driving vehicles and robotaxis that will one day operate without a human behind the wheel, coping with inclement weather is a daunting challenge. It’s one of the reasons that developers of self-driving vehicles are using sophisticated weather data to help them operate more safely.

Alphabet’s self-driving unit Waymo, for example, is preparing to expand its self-driving service to San Francisco, California, and Phoenix, Arizona, before rolling out to new cities in the United States. In a recent blog post, Waymo shared some of the ways its self-driving robotaxi vehicles deal with rain, fog, and other bad weather conditions.

In San Francisco, where Waymo recently launched its robotaxis service, the fog that rolls in each evening from the Pacific Ocean is something the company’s self-driving vehicles need to be able to handle.

An autonomous vehicle relies on perception sensors, much like a human driver uses their eyes and ears to navigate safely. However, bad weather can be difficult for humans and can impact self-driving systems.

For example, condensation can corrupt sensor data and fogging up a camera lens can degrade its performance. Additionally, water droplets, road grime and ice can accumulate on vehicle sensor surfaces and degrade their sensing capabilities, while wet roads can become more slippery. For Waymo’s software-based autonomous vehicles, overcoming these challenges is vital to the safe operation of its commercial robotaxis that will transport passengers.

Waymo’s latest hardware, including cameras, radar and lidar, uses raindrops on vehicle windows to classify various weather conditions. This data is then combined with high quality terrain data from weather visibility sensors. All of this information is used by Waymo to generate a quantitative weather visibility metric.

Waymo says this metric is used to perform analysis of the weather around one of its autonomous vehicles in real time, essentially making each Waymo robotaxi vehicle a self-contained “mobile weather station”.

Although ground-based weather stations and satellites provide meteorologists with accurate data to make fairly accurate weather forecasts, Waymo says these high-tech weather tools and datasets often lack the precision and specificity to accurately reflect conditions. in which its vehicles drive.

For example, some of this Doppler radar weather data is measured at altitudes of approximately 3,500 feet and up to 9,000 feet. This may not correspond to current real-time weather conditions in the field where Waymo’s self-driving vehicles operate. Additionally, many weather stations are located near major airports and are not always the best and ideal source of real-time local weather information.

San Francisco’s climate can vary greatly from neighborhood to neighborhood, as locals can attest. Residents on the east side can enjoy the sun while the city’s Outer Sunset neighborhood can experience dense fog at the same time of day.

While remote weather data from satellites or radar can help fill in the gaps of local weather stations, they also don’t directly detect weather conditions on the ground where Waymo’s vehicles operate, especially when it’s cloudy. , according to Waymo. Thus, the company’s fleet of autonomous vehicles serve as mobile weather stations combining millions of data points to more accurately predict weather conditions on the ground for the entire fleet.


Doppler radar does not always reflect actual weather conditions on the ground, where autonomous vehicles operate.

This additional layer of weather data is used for Waymo’s ridesharing services in San Francisco and Phoenix, and the company said it will create similar weather maps for other cities as it expands its ridesharing service. autonomous Waymo One.

Waymo even created the first-of-its-kind “fog map” for San Francisco that offers unprecedented spatio-temporal resolution. It allows the company’s autonomous fleets to track the progress of coastal fogs as they arrive daily from the Pacific Ocean and burn away as the sun rises later in the morning.

Waymo’s autonomous weather detection vehicles can even detect drizzle and light rain that is invisible to the National Weather Service‘s local Doppler weather radar. This detailed weather data allows Waymo to locate where weather conditions are beginning to deteriorate or where they are improving.

This detailed understanding of current weather and its impact on Waymo’s self-driving technology stack presents additional opportunities for designing robot fleets that will frequently encounter varied weather conditions. For one, Waymo engineers can fine-tune the performance of the Waymo Driver in various weather conditions using various data augmentations and simulated environments representing the seasons of the year.

Since Waymo uses the same suite of sensors to navigate and autonomously collect weather data, the company will create a vast dataset that can be used to further enhance its weather simulation capabilities.

Waymo’s understanding of the weather is also used by its autonomous freight delivery company Waymo Via. Going forward, it will be used to help Waymo Via delivery partners and their customers by providing more accurate updates regarding their freight in transit, especially for longer haul freight deliveries over longer distances that cover multiple states and geographic locations.


About Jefferey G. Cannon

Check Also

Gorgeous weekend weather across Arizona

PHOENIX (3TV/CBS 5) — Temperatures are starting to get chilly across Arizona, with valley neighborhoods …