Skip to content

September 25, 2018

Microsoft and Shell build A.I. into gas stations to help spot smokers

by John_A

The last thing you want to see when you pull into a gas station is some doofus lighting up a smoke.

Whether they missed the warning notices or, perhaps, the science class back at high school about open flames and flammable vapor is, in that moment at least, largely immaterial.

As for your own course of action upon seeing such reckless behavior, you can either put your foot down and hightail it out of there before the whole place goes up, or yell at the smoker to put it the hell out.

Tackling the very same issue, Shell has been working with Microsoft on a solution that aims to make all future visits to gas stations stress-free, at least in terms of potential explosive activity. The system uses Microsoft’s Azure IoT Edge cloud intelligence system to quickly identify and deal with smokers at a gas station, and it’s already being tested at two Shell stations in Thailand and Singapore.

It works like this: High-tech cameras positioned around the gas station filter the footage on site to identify behavior that suggests someone is lighting up, or already smoking.

Images that appear to show such behavior are then automatically uploaded to the Microsoft Azure cloud, which can power more sophisticated deep learning artificial intelligence (A.I.) models to confirm whether the person is actually smoking. If so, an alert is sent immediately to the station manager who can then shut down the pump before anything potentially cataclysmic happens. The system presumably could also be fully automated and configured to shut down the pump without the manager having to do it manually, with an audible warning given to the smoker via a speaker in the pump. Taking it to the extreme, the setup could even blast the perpetrator with foam from a fire extinguisher incorporated into the pump.

The entire process, from identification to shutdown, can take place in a matter of seconds. Shell said that this is because so much of the initial data is processed by on-site computers rather than sending everything to the cloud for processing — a feature of Azure IoT Edge. In other words, only the important data — in this case images that appear to show someone smoking or about to smoke — is sent to the cloud, a procedure that helps to speed up analysis and response time.

“Each of our retail locations has maybe six cameras and captures something in the region of 200 megabytes per second of data,” said Daniel Jeavons, Shell’s general manager for data science. “If you try to load all that into the cloud, that quickly becomes vastly unmanageable at scale. The intelligent edge allows us to be selective about the data we pass up to the cloud.”

As Microsoft explains on its A.I. blog, the intelligent computer vision tools could be used in a range of industries to automatically detect dangerous behaviors or conditions, for example, “it could be deployed on construction projects to flag when employees aren’t wearing proper safety equipment or to inspect equipment sitting on the seafloor thousands of feet underwater.”

Editors’ Recommendations

  • Ford recalls 2 million of its popular F-150 trucks due to fire risk
  • OneDrive leans on A.I. to simplify searches for multimedia files
  • Armormax’s AWD Dodge Charger SRT Hellcat may be the ultimate cop car
  • Get your Sagan on with 60 awe-inspiring photos of the final frontier
  • Don’t be fooled — this automated system sneakily manipulates video content



Advertisements
Read more from News

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Note: HTML is allowed. Your email address will never be published.

Subscribe to comments

%d bloggers like this: