<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1178143905890834&amp;ev=PageView&amp;noscript=1">

    I spy with my computer vision...too many cars in the drive-thru queue!

    Posted by Sabari Ayyappan M Mar 24, 2021 3:09:39 AM

    When Skellam AI’s brightest engineers tossed around in their beds, pondering about improving the business metrics of our QSR customers, they had a brain wave! They started to count…no not sheeps...but cars!

    Among the critical metrics of a QSR drive-through, the average service time ranks #2, only second to store location/layout. It is no surprise that QSR brands compete on a margin of seconds, as the speed of service has become the name of the game. However, QSR magazine reports that there is a gradual increase in the wait time over the last couple of years; in some cases up to 20 seconds slower in 2019.Image 1

     

    In 2020, a difference of 30 seconds in servicing resulted in over $32,000 of lost business per store in the U.S.

    Image 2

    When the stakes are so high, we realized that our QSR clients could do with an extra pair of eyes to keep the queues short. Skellam AI’s computer vision based solution can track the number of cars in queue, process the difference between expected service time &  the actual service time in order to optimize operations and enhance CX.

    Computer Vision

    Here is what you are seeing in the above video of one of our earliest PoCs. This is a single CCTV video recording with augmented computer vision. You see vehicles pulling up into the drive-through. 

    1. The algorithm tracks the cars and assigns a unique ID to each one
    2. In this version of the solution, you will notice that the ID is not always sequential - rather, it is assigned by order of identification 
    3. The car count gets computed in real-time depending upon the number of cars identified at a given point in time.
    4. Do note, in this video only the serving time is calculated; not the wait time
    5. The serving time is measured & displayed only for vehicle # 2. - as it is the only vehicle that enters and exits the serving area - others directly move to the pick up point
    The applications in drive-through are almost endless. We have put together the top applications that we are working on with our QSR clients.
     
    1. Switching pre-sell digital displays to quicker-to-serve combos during high traffic times
    2. Providing best discounts on quicker-to-make orders during peak service times
    3. Displaying offers on pre-sell menu for higher order value options during slack times 
    4. Allowing app users to pick the quickest drive-through vs the nearest (in real-time)
    5. Optimizing employee scheduling (shifts rostering) at drive-throughs
    6. Running campaigns to even out the servicing time & idle time; think happy hours!
    7. Making quick assessment for investments such as order confirmation boards, better speaker systems, layout redesign, digital pre-sell boards, customer service training etc.

    Do you have any other use case in mind? We would love to hear from you. Write to sabari@skellam.ai if you’d like to know more about our computer vision based solutions for quick service restaurants. 

    Our engineers continue to make the AI count cars so that our QSR clients don’t lose sleep over average service time.

     

    Leave a reply
    logo-dark-1_optimize

    Skellam is a team of experts in AI, data science and product development that develops custom AI solutions to solve complex business problems.

    Drop Us A Line

    3600 136th PL SE #300
    Bellevue, WA