Sign In  |  Register  |  About Burlingame  |  Contact Us

Burlingame, CA
September 01, 2020 10:18am
7-Day Forecast | Traffic
  • Search Hotels in Burlingame

  • ROOMS:

AI is teaching the Ford Mustang Mach-E how to drive

Ford has been using artificial intelligence to improve the operation of its semi-autonomous BlueCruise advanced driver assistance system and add new functionality to the feature.

The Ford Mustang Mach-E has been going to driving school.

The electric SUV is one of the models that is available with Ford’s BlueCruise semi-autonomous adaptive cruise control system.

The feature is similar to GM’s Super Cruise and allows drivers to take their hands off the wheel and feet off the pedals while the car controls its own speed and steers within a highway lane.

Facial recognition technology makes sure that the drivers are keeping their eyes on the road and are ready to take back control in case of emergency.


Ford's BlueCruise will deactivate if they are not, and the car will bring itself to a controlled stop with the hazard lights on if a driver fails to respond.

At least that is all it could do when it launched a year ago. 

The 2023 Mustang Mach-E comes with the latest BlueCruise 1.2 software, which has added functionality and works better overall.

Ford has been using artificial intelligence to improve the system’s capabilities through machine learning.

Along with its in-house testing, Ford collects data from owners who have opted in to share real world information from their vehicles, which includes video clips and radar records.

Ford has sold nearly 200,000 vehicles with BlueCruise to date, and customers have used it for more than 64 million miles combined, which has provided it with a huge amount of data to work with.

"A fairly large number of customers do decide that they want to share data with us," Sammy Omari, Ford’s executive director for advanced driver assist system technologies, told Fox News Digital.

"We are very surgical about what we’re allowed to collect and not just about how we collect it, but how we handle that data internally," he said.


AI software scrubs the data of personal information, blurs license plates in the images and crops them to just the parts it needs.

"We’re not really interested in anything, any data that could lead us to what customers used the car, like, for example, things like the vehicle identification number," Omari said.

The vehicles send reports to Ford over cellular or Wi-Fi connections whenever BlueCruise disengages, so that it can see what happened.

"Every time a customer has a disconnect out of hands-free to hands-on, we do collect that little bit of data, and we automatically analyze the data," Omari said.

"We want to know how our customers are using our product, where does the product sometimes fail. It’s not perfect."

However, it is getting there. Compared to when Fox News Digital tested the original version last year, BlueCruise can now stay engaged for much longer before asking for help.

"For time between disconnects we improved almost 2x, so it’s kind of dramatic," Omari said.


The algorithm learns in part by looking at video shot in different environmental and lighting conditions. 

This gives it a broader understanding of what roads and lane markers look like across the 130,000 miles of controlled access highways in the U.S. and Canada that have been certified for it to operate on so far.

It also took a few notes from the drivers. The system realized that drivers would often take the wheel and move over a little bit within their lane when a large truck was next to them, so BlueCruise 1.2 has been programmed to do that automatically when it is operating.

"We saw that in our customer data over and over and over again," Omari said.

It is also now able to check for other vehicles and change lanes by itself when the driver hits the turn signal, and generally deals with traffic in a more natural way. Additional capabilities will be added down the road, but Ford hasn't revealed them just yet.

Of course, while AI does a lot of the heavy lifting developing the software, it is ultimately up to the humans on Omari's team to add the final touches and validate it. The cars cannot modify the software onboard by themselves.

"Machine learning engineers are a very rare commodity and a very expensive commodity," Omari explained.

"For them to operate at scale, at Ford scale, we really want them to focus on small bits and pieces where they can add value that we can’t automate."

Data & News supplied by
Stock quotes supplied by Barchart
Quotes delayed at least 20 minutes.
By accessing this page, you agree to the following
Privacy Policy and Terms and Conditions.
Copyright © 2010-2020 & California Media Partners, LLC. All rights reserved.