Here are some use cases of LiDAR being used in retail to optimize operations and create better experiences for customers.
Customers walk around stores, browse displays, walk down aisles, and stop to consider products. From the outside, it may not appear there is much to glean from these behaviors. But with these everyday interactions, customers are telling a lot about their interests and how well they are being served. For retailers, the challenge is capturing the data that will help them understand what customers are telling them.“Remote sensing” can detect the physical characteristics of an environment, including the movement of individual. The data collected can help us understand customer behavior and optimize retail experiences. LiDAR (Light Detection and Ranging) is an especially promising part of this technology that provides an extra level of dimension that 2D sensors cannot.
LiDAR generates pulses of light using a laser and measures the time it takes for that light to be reflected back to the device. By measuring the reflection time at many points, we are able to generate 3D representations of objects in an environment.LiDAR has been adopted by the autonomous vehicles industry because it provides computer vision models with a new dimension of information without losing any valuable information from 2D. Done right, LiDAR sensors help perception models understand their surroundings and make sense of how they can move through them.While LiDAR is often spoken of in the context of the AV industry, there are a slew of applications that are emerging in retail — many of which circumvent obstacles commonly seen by automobile and parts manufacturers.While LiDAR data captured by autonomous vehicles is notoriously unreliable in unpredictable weather conditions, these challenges don’t apply for retail: there are no drops of rain confusing the sensor or calibration noise caused by moving sensor.
Indoor settings with more stable lighting conditions and fixed camera locations naturally lend themselves to cleaner, more reliable data. What’s more, retail as an industry can benefit from the technical improvements and innovation that the autonomous vehicle industry has made to LiDAR out of necessity. For example, AV has paved the way for cheaper LiDAR sensors, and annotation partners dedicated to providing cheap and accurate annotations.Read on to learn about some retail LiDAR use cases, and how retailers are already using this technology to understand more about their customers’ behavior to optimize their processes and create more delightful experiences.
As stores reopened during the pandemic, retailers had a new challenge: ensuring customers were safe by following social distancing protocols. Of course, retailers could hire someone to monitor the spacing of individuals, but that would be cost prohibitive and error-prone. Instead, some turned to LiDAR to detect the location and movement of shoppers relative to each other. LiDAR is especially well suited for this, as it uses true pose information, which is much more reliable than having to estimate distance from a set of 2D images.As stores reopened during the pandemic, retailers had a new challenge: ensuring customers were safe by following social distancing protocols. Of course, retailers could hire someone to monitor the spacing of individuals, but that would be cost prohibitive and error-prone. Instead, some turned to LiDAR to detect the location and movement of shoppers relative to each other. LiDAR is especially well suited for this, as it uses true pose information, which is much more reliable than having to estimate distance from a set of 2D images.Home Bargains used LiDAR to control the cost of complying with public health mandates. The technology worked particularly well compared to other types of surveillance technology — one benefit is that LiDAR is less susceptible to noise introduced by sunlight coming and going during the day.Retailers are also turning to LiDAR as they adapt to pandemic-inspired changes to business operations. For example, one retailer shifted to shipping more orders and used LiDAR to help monitor packing operations to ensure the quality of the process.
Lowes, a home improvement retailer, uses LiDAR in the "Measure Your Space" feature of its iOS app, which allows customers to scan rooms at home to help with their home improvement projects.The app guides customers through the process of scanning their room, and then automatically generates a floor plan, room measurements, and even a personalized estimate. Applications like this show just how accessible LiDAR technology is to consumers, opening up a world of possible use cases for forward-thinking retailers. As Seemantini Godbole, CIO at Lowes, put it:
“We see a future in which the devices customers already own can sense, understand and compile information about their home, putting it in their hands the moment they need it.“
Online retailers have no shortage of data about their customers thanks to a seemingly endless supply of clickstream data. Brick and mortar retailers do not have the ability to capture that kind of data, but they may have something better. Video analytic solutions bring clickstream data to in-person shopping. With LiDAR technologies, retailers can calculate dwell times of customers and detect which product lines are drawing the most customer attention. Imagine for a minute the customer intelligence you could unlock if you could combine 3D and 2D sensor data (such as LiDAR and video) to bring deep insights about a shopper’s movement/pose information and the attributes of the objects they interact with. Combined with video, there is the potential to produce a wealth of insights about customer location and movements:
The amount of shopping intelligence that could be produced and analyzed thanks to LiDAR and video is almost boundless.
LiDAR provides the raw data for a more complex analysis of customer behavior. For example, retailers can track individuals as they move through a store. This allows analysts to understand which isles are visited most, which products attract the most attention, and what paths through the store customers take.This is crucial to identifying high-traffic and low-traffic areas within a store. This kind of data is particularly useful in understanding the impact of store location and product selection on customer behavior. It can also be leveraged to alert managers when checkout lines are growing.
For these more nuanced insights into customer behavior, the full value of the data collected by LiDAR is realized by applying advanced data analytics and computer vision techniques. Computer vision systems, in turn, require machine learning models trained with uncompromising high-quality annotated data — for 3D data, accurate ground truth annotations are needed to ensure the reliable performance of computer vision and perception systems, and Sama can help.
Sama provides ML Professionals and AI team Leads with an indispensable solution for Computer Vision data labeling.
