How Retailers Can Reduce Inventory Shrinkage With Automated Asset Tracking With Calamp and AWS

The importance of asset tracking

Every day, Amazon.com receives millions of shipments from suppliers. A purchase order from Amazon to a supplier typically includes a large quantity of shipping containers, or assets, that will be delivered to various Amazon warehouses at different times. To determine if items are lost or misplaced in the large, complicated shipments, Amazon uses inventory shrinkage as a metric to measure the proportion of assets that suppliers claim to have shipped versus the quantity that was delivered to Amazon fulfillment centers.

By measuring inventory shrinkage, both Amazon and suppliers can check if purchase orders are fulfilled in full, identify theft-prone areas on transportation routes, and detect whether products might have been shipped to the wrong Amazon fulfillment center.

Many retailers, who don’t have an automated asset tracking system, have to manually count packages or containers and compare the assets received in a shipment to purchase orders. This labor-intensive process makes it difficult to detect when products have been shipped to the wrong fulfillment location or if items have been lost or misplaced during transit.

Amazon used Amazon Web Services (AWS) to build an automated asset tracking platform. The solution uses Telematics Cloud solution from CalAmp, an AWS Partner, to connect devices that monitor and visualize packages in shipments. In this blog, I will describe a complete Internet of Things (IoT)–based asset tracking solution featuring AWS and CalAmp technologies. I’ll also share a reference architecture and several use cases.

Roles in asset tracking process

Before I dive into the solution, I’ll set the stage by explaining the three key roles involved in the asset tracking process:

• Asset tracking administrator—This person monitors purchase orders, shipments, and assets in Ideally, the administrator role needs to visualize the monitoring metrics to detect inventory shrinkage.
• Fulfillment center operators—In this role, someone manages the inbound assets as they are delivered to fulfillment centers, and they oversee the unloading, counting, and recording of items in the delivery compared to the purchase order. Ideally, the role needs an automated approach to replace physical asset counting and recording.
• Product suppliers—This function manages outbound assets that are shipped out of factories or warehouses and are transported to fulfillment centers. The responsible person oversees the counting and loading of assets onto transportation vehicles and informs Amazon that a certain quantity of items has been shipped. Ideally, the role needs closed-loop notification from Amazon that all assets have been received in good condition at the fulfillment centers.

High-level overview of asset tracking solution

To accurately track assets and automate the tracking process, you can use IoT sensors to tag packages and collect telemetry location data to indicate whether assets in transit enter or leave specific locations. Ideally, a technology solution should

• automatically count, weigh, or measure assets delivered by suppliers to Amazon fulfillment centers;
• compare purchase orders to delivered quantities to detect inventory shrinkage;
• notify fulfillment center operators to take actions when missing assets are detected;
• analyze asset tracking data to identify missing items or locate incorrect deliveries; and
• visualize relationships among purchase orders, shipments, and assets to uncover insights about inventory shrinkage.

Now that I’ve set the stage with a high-level overview of asset tracking, I’ll walk through the complete AWS and CalAmp approach.

CalAmp IoT asset tracking solution powered by AWS

CalAmp’s IoT-based solution uses gateways and tags to track assets and distribute the information to different system users. The data can be translated and merged with business logic to provide critical insights about assets in transit, or the raw data can be loaded into a repository for future analysis.

Figure 1

The figure 1 image shows that a supplier received a purchase order from Amazon with the asset amount and loaded the assets onto a truck for transit. Once the truck crossed the CalAmp gateway at the supplier warehouse, the ID and state data from tagged containers was sent to the backend AWS asset tracking solution. The solution automatically created a shipment record and attached it to the purchase order, indicating that the assets left the supplier warehouse and were in transit to the fulfillment centers. It’s important to note that assets moved within the supplier warehouse that do not cross the CalAmp gateway will not initiate shipment records to change the asset state.

Figure 1. Assets leave supplier warehouse

Figure 2

In this image, the truck has arrived at the fulfillment center and crossed the CalAmp tracking gateway. The IoT sensors automatically sent asset ID and state data to the backend asset tracking solution. The solution retrieved the tracking information from when the assets left the supplier warehouse and matched the arrival information to identify and record whether assets have arrived safely, were lost, or have gone to the wrong Amazon fulfillment center. If all items arrived as expected, the solution will record the assets and the shipment as complete. If items were lost or misplaced, the application will send notifications about the lost or misplaced assets to the administrators.

Figure 2. Assets arrive at Amazon fulfillment center

Asset tracking IoT edge devices

The solution uses two types of edge devices from CalAmp: gateways and tags. Gateways are installed on the warehouse gates of product suppliers and Amazon fulfillment centers, and they connect to CalAmp’s Telematics Cloud solution to track whether asset tags cross the entrance and exit gates. The asset tags are Bluetooth Low Energy (BLE) sensors that provide proximity location. All dependencies for the gateway and tag sensors as well as the tagged assets are stored in the solution.

Asset tag device lifecycle

When the asset tracking administrator receives the CalAmp gateways, they install the devices on the warehouse gates and connect them to the CalAmp Telematics Cloud through the internet. By default, the gateways are set to “idle.” Within the solution, the asset tracking administrator can assign or reassign gateway sensors to suppliers and fulfillment centers with just one click, and when the devices are uninstalled, the administrator can recycle the gateways. The gateways have two states: “idle” and “assigned.”

Figure 3 shows:

1. The asset tracking administrator assigns tags to a product supplier and ships the tags to the supplier. If the tags are no longer needed, the administrator can recycle them and use them with another supplier.
2. The product supplier pastes the tags to the outer packaging of the assets that will be transported to the fulfillment centers. In the application, the supplier can scan a tag and input information specific to the products associated with the tag, such as purchase order number and fulfillment center ID. After that, the tag state will be automatically changed to “associated.”
3. The product supplier loads the tagged assets onto a truck for transport. When the truck crosses the installed sensor gateways as it leaves the supplier location, the tag state will automatically change to “in transit.”
4. When the truck arrives at an Amazon fulfillment center and passes another installed CalAmp gateway, the asset tags in the truck will automatically change to a “completed” state.
5. The solution follows predefined rules to identify and notify the administrator about lost or misplaced assets. Meanwhile, for assets that arrived at the fulfillment center with no issues, the solution automatically converts the tag state to “assigned.” These tags are ready for reuse, so the fulfillment center operator can detach the tags from the asset packaging and ship them back to the supplier or ship the containers with the tags back to the supplier.
6. From within the solution, the asset tracking administrator monitors tag use among all suppliers, and they assign and recycle tags as necessary with each supplier.

Figure 3. Lifecycle of tags in the solution

Asset tracking platform

The CalAmp asset tracking platform is built on AWS, and it processes telematics data ingested from the CalAmp IoT edge gateways and asset tags to provide asset tracking administrators, fulfillment center operators, and product suppliers with the functionality and data that they need to monitor the movement of goods from suppliers to fulfillment centers, locate misplaced items, and detect lost assets with inventory shrinkage analytics and visualizations.

Figure 4 depicts the platform architecture.

1. The CalAmp gateway detects asset tags when the tags enter an area covered by the gateway.
2. The gateway sends telemetry data about the tags to the CalAmp Telematics Cloud platform.
3. The CalAmp Telematics Cloud platform converts the telemetry data to a JSON format and sends the data to the asset tracking platform through Amazon Simple Queue Service (Amazon SQS), a fully managed message queuing service.
4. When Amazon SQS receives the JSON data in the asset tracking platform, an AWS Lambda function is initiated. AWS Lambda is a serverless, event-driven compute service.
5. The AWS Lambda function purifies the data and sends the purified data to Amazon EventBridge, a serverless event bus that makes it easier to build event-driven applications at scale.
6. Amazon EventBridge transmits the data to the downstream services.
7. One AWS Lambda function receives, parses, and writes the tag data (tag state, purchase order number, and others) to Amazon DynamoDB Amazon DynamoDB is a fully managed, serverless, key-value NoSQL database designed to run high-performance applications at scale.
8. Another AWS Lambda function also receives the data sent by Amazon EventBridge.
9. The AWS Lambda function retrieves the information about shipment records, fulfillment centers, and suppliers from Amazon DynamoDB and, with the data from Amazon EventBridge, the AWS Lambda function identifies lost or misplaced tags. If lost and misplaced tags are detected, the AWS Lambda function invokes a notification from Amazon Simple Notification Service (Amazon SNS), a fully managed messaging service.
10. Amazon SNS notifies the asset tracking administrator about the lost or misplaced assets in the shipments along with the associated purchase order information.
11. The schedulers for cron jobs are defined in the event rules in Amazon CloudWatch, a monitoring and observability service. The job initiates the AWS Lambda function to check if lost or misplaced tag notifications have been ignored.
12. The asset tracking administrators, fulfillment center operators, and product suppliers access the solution by logging into a web portal hosted on Amazon Simple Storage Service (Amazon S3), an object storage service built to retrieve any amount of data from anywhere. The users are authenticated and authorized to use the system by Amazon Cognito, a service that lets you control access to AWS resources. Once users are logged into the system, they can perform their asset tracking functions, such as assigning tags to suppliers and monitoring shipments in transit.
13. The web portal loads data requested by the different users and processes each user’s input through AWS AppSync, a fully managed service to develop GraphQL APIs.
14. AWS AppSync updates data stored in the Amazon DynamoDB tables.
15. The AWS Lambda function maintains the raw data and removes only null and error data, converts the data into JSON, and redirects the data to Amazon Kinesis Data Firehose, an extract, transform, load (ETL) service that reliably captures, transforms, and delivers streaming data to data lakes, data stores, and analytics services.
16. Amazon Kinesis Data Firehose stores the JSON files in Amazon S3 buckets.
17. Crawlers created in AWS Glue—a serverless data integration service that makes it easy to discover, prepare, and combine data—grab data from the Amazon S3 buckets and form databases.
18. The crawler data is extracted, transformed, loaded, and stored in Amazon DynamoDB tables and also form databases in AWS Glue.
19. Amazon Athena, an interactive query service to analyze data in Amazon S3 using standard SQL, acts as data sources of Amazon QuickSight to provide visualization data. Amazon QuickSight is a cloud-native, serverless, business intelligence solution that allows you to understand data by asking questions in natural language, explore interactive dashboards, or look for patterns and outliers with machine learning.
20. The asset tracking platform creates many datasets from Amazon Athena for analysis in Amazon QuickSight and composes inventory shrinkage dashboards so that users can visualize gateways and tag states and shipment statuses.
21. Figures 5–9 show examples of different views of interactive dashboards that asset tracking administrators can see, filter, and sort in Amazon QuickSight.

Figure 4. Asset tracking platform architecture

Figure 5. Example of asset tracking platform dashboard

Figure 6. Example of asset tracking platform dashboard

Figure 7. Example of asset tracking platform dashboard

Figure 8. Example of asset tracking platform dashboard

Figure 9. Example of asset tracking platform dashboard

Contact AWS today

If you’d like to know more about the AWS and CalAmp approach to streamlining and automating asset tracking and inventory shrinkage processes, contact your AWS account team today.

To learn more about AWS IoT solutions, please refer to our IoT solutions page.