Increased connectivity and smart displays can help deliver new forms of intelligence about how shoppers interact with physical retail spaces. Detailed information about shopper activity is beneficial to both business owners and consumers, helping businesses maximize profits by making smarter shop floor layout choices and leading to a more effortless and enjoyable shopping experience for consumers.
We've combined Texas Instruments hardware with Temboo's cloud platform to build an IoT solution for physical retail spaces that passively collects data and makes it available for analysis via custom mobile apps.
Android device
TI SensorTag
Parse is a cloud-based platform that helps you focus on creating a great end user experience, without having to worrying about server maintenance and complex infrastructure. Putting your data in Parse gives you a natural extension point for building apps on top of that data - in this case, apps that report shop floor activity to multiple distributed stakeholders.
Twilio exposes a globally available cloud API that developers can interact with to build intelligent and complex communications systems that use voice, VoIP, and SMS text-messaging.
To get started building this application, we set up an Android device so that it could communicate with Parse and add rows of data to our Parse data store. To do this, we used the Parse > Objects > Create Object Choreo to generate code that uses Temboo's Android SDK, and passed it some test sensor data to store in Parse. With our Android app up and running, we set up our TI SensorTag to generate real sensor data that was be transmitted from the SensorTag to the nearby Android device via Bluetooth Low Energy.
When a SensorTag device is turned on, it generates data with all six of its sensors. For this particular application, we were interested only in data from the accelerometer and the gyroscope, which allow us to determine activity around retail displays.
The SensorTag comes with an open source mobile app that reads and displays data generated by the onboard sensors on iOS and Android devices. Augmenting this app with the Temboo Android SDK code that we generated earlier enabled us to send the sensor data to Parse via the Android device's internet connection.
Once our data is in Parse, we can use any of our SDKs to extract the data and build native or web-based mobile apps with which interested parties can consume the store traffic data.
Additionally, we used our Twilio Choreos to trigger automated phone calls to store managers whenever notable events were detected e.g., a display was low on stock.
Choosing a data store often feels like a big commitment, but Temboo makes it easy to switch out one data store for any of the many other data storage options we offer. For example, we could easily switch out Parse for a MySQL database running on Amazon's cloud platform by using our Amazon RDS Choreos.
As well as consuming the data in report format, stakeholders want real-time alerts when certain retail store activity thresholds are hit. We integrated phone alerts by using our Twilio Choreos, but we could easily swap these out for another telephony provider e.g. Nexmo. Additionally, we could add different types of alerts, like status emails via Gmail or customer support tickets via Zendesk.
Aside from building on top of what we've got, or easily switching out the web-based services we've selected, we could also reimplement this project using Temboo's support for Arduino hardware.
Contact us at hey@temboo.com