Blog post

Computer Vision AI Well Positioned for Post-Pandemic Shopping

By Anthony J. Bradley | May 07, 2020 | 2 Comments

Computer vision technology is poised to accelerate new post pandemic shopping experiences. To say that the grocery shopping experience in the US (and other countries) has changed is a gross understatement. The other day I went into a supermarket to pick up a few items and it was eerie to see people walking around in masks.

One guy thought it would be hilarious to wear his son’s Darth Vader mask into the grocery store. OK, that guy was me. Nobody found it funny. Fundamentally, there is something scary about a person wearing a mask and when everybody is wearing a mask it is a surreal and discomforting experience.

My wife won’t go into the grocery store at all. I’ve spent years trying to get her to order online and it took a global pandemic to change her behaviors. But she exclusively orders for delivery or curbside pick up. This pandemic is changing behaviors.

Don’t get me wrong. The delivery and curbside experiences also leave much to be desired. It makes sense. Peapod and other grocery delivery services were not set up for this scale and so getting groceries delivered is a little like getting a reservation at that elusive Michelin star restaurant. My wife will get up at 2 AM in the morning to try to schedule on newly opened time slots. Whole foods (powered by Amazon) is just as challenging. It takes a lot of time and effort to order groceries only to have about half of them show up. Still, she is grateful for the options and not thrilled with those grocers who have decided not to adopt new practices.

Most grocery stores were also slow to embrace effective curbside pick up. HEB is probably the exception to the rule where they turned almost half their parking lot into pick up spaces. And curbside pick up is the only real option for local grocery stores who have little to no web presence. E-commerce is not an option for them.

How much of these online ordering and curbside pick up behaviors will endure after the pandemic is hard to predict. Much of it depends on how long The pandemic lasts. but it is safe to say that there’s a significant chance that we will experience a new normal rather than a return to the way things were.

Computer vision and video analytics technologies are ready to empower new safer and more efficient shopping experiences for both buyers and employees. Nick Engelbrecht, my lead analyst for computer vision technology, just published a great note entitled “Top Advanced Computer Vision Use Cases for Retail“ (only available to Gartner clients).

In this research he highlights the leading edge of retail use of computer vision and video analytics. He dives into examinations of use cases such as grab and go, shelf vigilance, behavior analysis and proactive security. 

  • Grab and go means just that. It provides a frictionless experience where computer vision recognizes what a shopper places in a cart or basket. Payment is automatically transacted and an e-receipt provided as the shopper exits the store. 
  • Shelf vigilance is a combination of computer vision, shelf sensors (perhaps even roaming robots) and machine learning to anticipate out-of-stock conditions, automated reordering and optimal placement. 
  • With behavior analysis shoppers are tracked (anonymously) to optimize shopper paths and minimize time-in-store. It also provides valuable information to store employees on fostering a more enjoyable experience.
  • Video analytics is enabling real-time assessment of potential threats. There is no lack of security camera footage in stores and malls. Using that video for quick and proactive security is where the challenge lies. Deep learning AI models are progressing rapidly at analyzing video to identify potentially threatening situations such as erratic behaviors, loitering, escalating conflict, crowding, etc. Real time monitoring and alerting allows security personnel to react quickly or even proactively to defuse potentially threatening situations.           

I believe that most of us sense that the COVID-19 pandemic is changing behaviors and accelerating digital business. I’ve got my teams of analysts focused on determining what behaviors will endure and the impact they will have on the advancement and adoption of emerging technologies. I’ll highlight some of these in future posts.

I’m very interested to hear your experiences and how you think emerging technologies will empower new post-pandemic realities. So please share.

Leave a Comment


  • pierre says:

    Hi Anthony, I work with a startup – and we are specialized in computer vision for the grocery.
    You mentioned numerous business cases, with different vectors of value for the retailers.
    What we see is the conversion of deep AI technologies on a single platform leveraging different AI libraries adapted to the business cases. From inventory to theft, retailers should find one size fit all.