ThingSpeak™ is an IoT analytics platform service from MathWorks®, the makers of MATLAB® and Simulink®. ThingSpeak allows you to aggregate, visualize, and analyze live data streams in the cloud. ThingSpeak provides instant visualizations of data posted by your devices or equipment. Execute MATLAB code in ThingSpeak, and perform online analysis and processing of the data as it comes in. ThingSpeak accelerates the development of proof-of-concept IoT systems, especially those that require analytics. You can build IoT systems without setting up servers or developing web software. For small- to medium-sized IoT systems, ThingSpeak provides a hosted solution that can be used in production.
ThingSpeak allows you to aggregate, visualize and analyze live data streams in the cloud. With ThingSpeak, your data is stored in channels. Each channel stores up to 8 fields of data. You can create as many channels as you need for your application.
You can use any Internet-connected device with ThingSpeak. When sending data from your devices or equipment, you can use native libraries for common embedded hardware prototyping platforms like Arduino®, ESP-8266, Particle and Raspberry Pi™. You can also send data to ThingSpeak from machines or local gateways using a REST API or an MQTT API. In addition, the following vendors have built integrations to ThingSpeak to make setup even easier:
If you are a Simulink user, you can use Simulink blocks in your models to write data to ThingSpeak. You can also and read data from ThingSpeak using embedded devices including iPhone, Android™, Arduino, Raspberry Pi and BeagleBone® Black.
ThingSpeak stores all the information you send it in one central location in the cloud, so you can easily access your data for online or offline analysis. Your private data is protected with an API key that you control. When you are logged in to your ThingSpeak account, you can use the web to securely download the data stored in the cloud. You can also programmatically read your data in CSV or JSON formats using a REST API call and the appropriate API key. Your devices can also read data from a ThingSpeak channel by subscribing to an MQTT topic. Import data from third-party web services including climate data from NOAA, public utility data from local utility providers, and stock and pricing data from financial providers. You can use that data together with the data you are collecting from your devices and equipment to investigate correlations and develop predictive algorithms.
MATLAB users can import data stored in ThingSpeak into the MATLAB desktop environment using the thingSpeakRead function.
ThingSpeak automatically charts the data that you send it, so you can remotely monitor your devices or equipment from anywhere. View your data from any web browser or mobile device. Share read-only views of your data with the clients and colleagues that you specify. Alternatively, you can use ThingSpeak to manage your data, and you can build your own front end for your clients and customers to log in to.
With a commercial ThingSpeak license, you can send data to ThingSpeak as fast as once every second. This not only enables near-real time monitoring of your devices, but it allows you to set up control loops from the cloud. For example, you could configure ThingSpeak to turn a light on when your motion sensor detects a person has walked into a room. For applications that require faster response times, the best practice is to have the control loop at the edge closer to the hardware.
With the MATLAB engine built into ThingSpeak, you can perform calibrations, develop analytics, and transform your IoT data. You can also use the MATLAB engine built into ThingSpeak to build custom charts. With a commercial ThingSpeak license, you can run MATLAB calculations that last up to 60 seconds. A commercial ThingSpeak license also enables you to use MATLAB Toolboxes for machine learning, signal processing, system identification, and more with ThingSpeak, provided you have a license for the toolbox.
Operationalize your analytics using the Time Control and React apps. With the Time Control app, you can schedule a computation to run once a day, once an hour, or as quickly as once every 5 minutes. The React App is used for condition monitoring. You can monitor the data coming in from your devices and set up an alert when the data indicates something may need attention. For example, you could configure ThingSpeak to send an email when the humidity on your plant floor exceeds a certain value. More broadly, your analyses can trigger events that push data from ThingSpeak to other web applications like Salesforce via REST APIs.
MATLAB offers thousands of robust and well-tested algorithms for statistics, machine learning, nonlinear optimization, system identification, image and video processing, signal processing, and control system design to use out-of-box or as a starting point in your IoT algorithm development. Use them out-of-box or as a starting point in your IoT algorithm development. MATLAB provides tools to build predictive algorithms based on machine learning or modeling-from-first principles.
Simulink supports a variety of hardware platforms for prototyping and building IoT systems. You can develop algorithms in Simulink and then deploy them on your embedded hardware using HDL and C code generation. Deploying algorithms on the smart device reduces the amount of processing required on the IoT platform and the amount of data that must be sent there.
With MATLAB and Simulink, you can use a single executable design – and deploy to cloud, edge, or embedded systems as needed. This allows you to perform trade-off studies and explore what needs to run where – leveraging the same algorithm or model. You can automatically generate code or software components for common algorithm or model deployment paths including C/C++, HDL, GPU, .Net, JAVA. This enables you to easily integrate your algorithm in the cloud, at the edge, or on an embedded system.