In the so-called 'new world' where the Internet of Things (IoT) is no longer a fancy buzzword, every object can be made ‘smart’. This means each object represents data and can share that data. With all the data these objects generate, IoT can be used to monitor and streamline industrial processes, logistics and more. This leads to better insights into those processes and as a result costs can be reduced. According to CSG International, of all the businesses who chose to implement IoT, 94% have already seen a return on their IoT investments.
Question - Query
When a company is distributing perishable goods to resellers around the world, it has control over the wares and the logistics, but only until the goods reach those resellers. What happens afterward? Are goods that are about to perish removed in a timely manner? Are its shelves not empty? If they aren’t, are they tidy? The company could send some people to physically go to the stores to make sure everything is up to snuff, as is currently the case. Considering the number of stores it works with (tens of thousands), this is a huge expense.
If the company wants to know what is happening once the products arrive at the stores, it would need its shelves to become smart! They would provide data concerning its contents. Everything that comes in or goes out of the shelf is monitored. You would not only see product movement, but also detailed information about the product, like the name, description, expiry date and more. Additionally, for marketing purposes and consumer behavior analytics, it would generate interesting data for studies and reports. The gathered information is being sent to and made available in the cloud. A dashboard provides an overview of the different (stores,) shelves along with their contents. Anomalies can trigger warnings on the dashboard.
To make the shelves smart, three key components are needed:
- RFID readers
- Products containing RFID tags (cheap RFID sticker)
- Smart device(s) (Raspberry Pi) to process data coming from the RFID readers
Data is sent to and stored on an SAP HANA database. Data is displayed on a dashboard (SAP Fiori Launchpad).
Proof of concept
All it takes is a hackathon of roughly a day and three Amista experts to transform a standard wooden shelf into a functional IoT hub. Nicely hidden RFID readers register all products (containing an RFID-tag) coming into and leaving the shelf. Raspberry Pis at the back of the shelf forward that data to an SAP HANA database. How? They each use Python scripts polling the RFID readers and Curl to send incoming data to the SAP cloud through the SAP IoT 4.0 API. All Pis are provisioned through Resin.io where they can be monitored and altered if needed.
In order to store information like stores, shelves, product specific info and the relation between them we use an SAP HANA database. This allows a seamless integration with SAP Cloud Platform. The SAP HANA database retrieves the results from the SAP IoT-services 4.0 and processes them to check if for example a product is still in or out of the shelf. Other calculations also happen inside the database. The expiration dates are checked and if a product has expired or will expire within 5 days, the data is prepared to that an alert will be shown in the SAP Fiori Launchpad.
The SAP Fiori Launchpad is deployed on SAP Cloud Platform and can communicate with the SAP HANA database this way. The launchpad has apps to show information like a list of shelves, what products are inside and a map with an overview of all the locations of the shelves. If a product will be removed from a shelf, the KPI tiles update themselves within a couple of seconds. The same goes for a product which is put on a shelf and for example would be expired, the launchpad will show an alert that an expired product is on a shelf.
Some examples of possible expansions could be the addition of a customer launchpad where a user can find more information about a specific product by scanning the RFID-tag with a smartphone or an app where the user can check if a product is in stock and where it is stored in the warehouse exactly. The authenticity of a product can also be checked, this way a counterfeit product can be easily identified.