Whether in a supermarket or department store, shelf and inventory management is the heart of any offline retail operation. Perfect retail shelving can enhance the layout design of a store and can create an inviting environment for customers. A strategically planned retail shelving can maximize customer engagement with the products on display and help increase revenue. In the past, store workers were tasked with enhancing shelving efficiency but, during the COVID-19 pandemic, retailers have had to think about ways to control social distancing, store and warehouse tidiness, and contactless processes for in-store shopping at the same time. To manage this, many have started relying on Artificial Intelligence (AI) in their stores.

In a study conducted by Retailwire, 77% of large retailers expressed they intended to create a clear robotics strategy in 2021. Those retailers listed shelf scanning (59%) and order picking (47%) as the top two types of robotic applications. Walmart began implementing robots in 2017, using the mobile robots to simplify routine work in stores and using machine vision to scan shelves and identify what products needed restocking to optimize inventory management.

In 2020, Walmart ended its contract with Bossa Nova Robotics, which provided the retailer with robots that scanned shelves for inventory. The Wall Street Journal reported that was likely due to the fact that the company hired extra workers to collect items for delivery or pick-up services during the COVID pandemic and discovered that they could also scan the shelves instead of the robots. It was also noted that the six-foot size of the robots Walmart was using could also have been obtrusive or intimidating to customers. This does not mean that the retail robot is done, but does show the importance of optimizing the robot design. Using Walmart as an example, we can determine that size and costs are important factors as is efficiency in terms of interaction with human workers and operating duration.

RSL10 Smart Shot Color Camera

The RSL10 Smart Shot Color Camera is a complete node-to-cloud solution platform for vision IoT applications that require always on, multi-year battery life with Bluetooth® Low Energy connectivity. Based on the RSL10 SIP and the ARX3A0 image sensor, the platform offers the same features as the RSL10 Smart Shot Camera in a smaller form factor and offers extended battery life and support for color and mono imaging.

Enabling edge-to-cloud Bluetooth Low Energy connectivity, the RSL10 SIP features an on−board antenna, RSL10 radio SoC, and all necessary passive components in a single, miniature package. This helps minimize the overall camera system size (50 mm circular form factor).

Alongside the impressive low power operation of the RSL10 SIP, all other system components on the camera platform have been selected to maximize the platform’s battery life. In the updated power management system, a coin-cell battery is used instead of a rechargeable battery, which greatly contributes to the ultra-low power consumption and cost of the bill of material.

The vision subsystem, which includes the image sensor and the image sensor processor, is placed on a separate power rail. This allows the controller (RSL10) to completely power off the vision subsystem for most of the camera operation duration and turn it on only to capture images when an event is triggered. The RSL10 SIP also remains in a deep sleep state most of the time and is only active when the camera is in advertising and connected mode when triggered.。

RSL10 Smart Shot Camera Power Consumption Measurements

  

The RSL10 Smart Shot Color Camera includes both color and mono IAS image sensor modules. Both are based on the ultra-low-power ARX3A0 image sensor with 1/10.3-inch and 0.3Mp fast rolling shutter imaging and 121° Diagonal Field of View (DFoV). By adding support for color imaging and wider FoV, the camera platform can enable new applications such as storage space monitoring, greenhouse monitoring or food quality monitoring.

The RSL10 Smart Shot Color Camera is supported by a mobile app, available on Android™ and iOS® platforms. Using the mobile app, developers can set trigger conditions, capture images and perform image analytics using the Amazon Rekognition™ (AWS) cloud platform.