♦ Title:
Smart Shelf using IoT - An Automated Intelligent Inventory System
♦ Field of invention:
• During festival seasons the operational efficiency of shelf can be improved by developing a smart shelf in the retail store.
• The intelligence given to the shelves help the retailers to maintain and track the
inventory in an efficient way. .
♦ Background of Invention:
In the retail industry, digital transformation is riding the next wave of innovation. Retailers are currently changing their in-store operations to keep up with technological improvements and increased competition from the on-line stores. Customers get visual and sensory experience in physical stores which cannot be replicated in e-commerce space.Retailers use shelf mainly to display and sell their products in offline.The inventory level of the products in the shelves is monitored by the store assistants manually and it is time consuming task.In the existing system, tracking of the products in the retail stores is very difficult.
♦ Summary of the Invention
The main objective of designing this "Smart Shelf using IoT - Automated Intelligent InventorySystem" is to help physical retailers compete with online merchants by engaging with customers and enhancing operational efficiency .The proposed system provides intelligence into on-the-shelf inventory levels without the labor cost of employees walking around the store. <* Unique Feature of this Invention:
The proposed Smart Shelves are intelligent inventory control system fitted
with weight sensors and Artificial Intelligence camera which are integrated with
NVIDIA Jetson Nano Developer Kit to improve the operational efficiency in the
Retail Store. The retailers collect real time and accurate inventory data from this
Smart Shelf, helps them to monitor the products quantity and periodical
popularity in an efficient way. / . (\ A ___—=5=-

♦ Proposed System:
The proposed system mainly consists of the following modules
1. Data & Image Acquisition
2. Intelligence Inventory Control Module Data & Image Acquisition:
Smart Shelf is designed with weight sensors. Weight sensors are used to find the total weight of the products in the shelf. The shelf is designed to have 50kg weights in the shelf. Artificial Intelligence camera is placed in front of the shelf helps to find which products are purchased by the customers and weight sensors gives the details about quantity of the products purchased by the customers.The weight sensors and Artificial Intelligence cameras are integrated with NVIDIA Jetson Developer Kit-B01. This receives input from weight sensors and AI cameras. Intelligence Inventory Control Module:
The Inventory control Module is developed to control the inventory level of the products in the shelf. The main advantage of this module is that the information is directly collected from the shelf and through AI cameras. The weight sensors provide us how much products are purchased in that shelf without any human intervention. The AI cameras capture the products at real time. This system helps us to process real-time images of the shelf. User can view in the interface updating to a low inventory warning threshold. ♦ Advantages:
1. This system is designed to track virtually any fast moving good in the retail store with extremely high accuracy rate.
2. It helps the retailers to track a wide variety of other product types.
3. The designed system provides a web-based interface to monitor the Shelf
\ in real time. It is set to create alerts at 50% percent and 20% thresholds.
4. These alerts will go to the designated staffs automatically or at set times, daily notifications can be sent in any format.
5. Finally, Smart Shelves enable real time and accurate management of inventory data to help retailers to optimize the in-store sales with the timely filling of stocks, review of items etc. and it also help physical retailers compete with qnlinevmerchants by engaging with customers in sophisticated manner. ,0)1

<* Brief description of the accompanying drawings
A better understanding of the present invention can be obtained from the description in detail accompanied by the following drawings, in which
• Fig. 1. shows the proposed Smart Shelf designed with weight sensors and Artificial Intelligence cameras and integrated with NVIDIA Jetson Developer Kit-BOl
• Fig.2. shows the Smart Shelf using IoT- Architecture diagram
• Fig.3. shows Inventory Control System
• Fig.4. shows weight sensor and is another term for a load cell. Weight sensor is a type of transducer, specifically a weight transducer. It converts an input mechanical force such as load, weight, tension, compression, or pressure into another physical variable, in this case, into an electrical output signal that can be measured, converted and standardized.
• Fig.5. shows the ESP8266, a low-cost Wi-Fi microchip, with built-in TCP/IP networking software, and microcontroller capability. This small module allows microcontrollers to connect to a Wi-Fi network and make simple TCP/IP connections using Hayes-style commands.
• Fig.6.shows the NVIDIA Jetson Nano Developer Kit.The Main components are
listed here
1. microSD card slot for main storage
2. 40-pin expansion header
3. Micro-USB port for 5V power input, or for Device Mode
4. Gigabit Ethernet port
5. USB 3.0 ports (x4)
6. HDMI output port
7. DisplayPort connector
8. DC Barrel jack for 5 V power input
9. MIP1 CSI-2 camera connectors
> NVIDIA Jetson Nano is an embedded system-on-module (SoM) and
developer kit from the NVIDIA Jetson family, including an integrated
128-core Maxwell GPU, quad-core ARM A57 64-bit CPU, 4GB
LPDDR4 memory, along with support for MIPI CSI-2 and PCIe Gen2
high-speed I/O. . O i^—-^—

> It is also a small powerful computer that lets you to run multiple neural networks in parallel for applications like Image Classification, Object Detection, Speech Processing, and Segmentation.
> But it can run in as little as 5 Watts.lt has great GPU(graphics processing unit) which makes it more popular for machine learning. Also used in gaming and CUDA based computations.
• Fig.7.shows JeVois-A33 Smart Machine Vision Camera.
> JeVois-Pro has an integrated 5-TOPS Neural Processing Unit (NPU), and works with optional 4-TOPS Google Coral Tensor Processing Unit (TPU) and 1-TOPS Intel Myriad-X Vector Processing Unit (VPU). Up to 13 TOPS inside (NPU + 2xTPU), and more using additional external USB accelerators.
> JeVois is an official partner of Google Coral, bringing you the latest capabilities in embedded deep learning, a solid edge computing development workflow based on Tensorflow-Lite, and amazing support.
> JeVois-A33 = video sensor + quad-core CPU + USB video + serial port, all in a tiny, self-contained package (28 cc or 1.7 cubic inches, 17 grams or 0.6 oz). Insert a microSD card loaded with the provided open-source computer vision algorithms (including OpenCV 4.x, TensorFlow, Caffe, Darknet, and many others), connect to your desktop, laptop, and/or Arduino, and give your projects the sense of sight immediately.
• Fig.8.shows EdiMax AC 1200 Wifi USB Adapter
> Wireless USB was a short-range,
highbandwidth wireless radio communication protocol created by the
Wireless USB Promoter Group which intended to increase the availability
of general USB-based technologies.