Platform for IoT Solution Development
Platform for IoT Solution Development
This project is ZYNK's platform for the rapid development of connected IoT apps. In short, the solution allows the user to collect, analyze, monitor, and react to data streams from connected devices. The data may also come from other enterprise systems.
With this project, we aimed at helping our clients build IoT solutions. Besides, our platform should permit fast, no-code development. We mainly targeted the product for smart buildings and smart cities. Additionally, we considered IoT systems in manufacturing, agriculture, and healthcare. Representing and connecting devices, measuring parameters, issuing alerts, and generating reports were all on our drawing board.
Three years of hard work led to our fully functional development platform for IoT and Big Data solutions. The solution features data harvesting, analysis, monitoring, control, and prediction services.
First, the user can create a visual representation of the IoT system on the platform. Next, they can manage the connected devices and the device configurations. In addition, the admin can handle the required software updates for the devices. Once configured, the system harvests data using mobile, IoT, or other intelligent devices. This data enters the system through communication protocols that support different types of data. Then, the user can analyze the data using the features implemented, including auto-machine learning.
Using a complex Event Processing System, our solution processes and correlates the events. Then, it converts them into system notifications, SMS, and mail alerts. The platform features user-friendly data monitoring with a broad set of widgets incorporated in the web application. These widgets include graphs, maps, tables, and video streaming.
The user defines events based on values for the metrics fed into the system. Then, for each event, they create intelligent alerts. For example, when the temperature in the server room reaches 20 degrees centigrade, the system administrator gets a warning via SMS.
Another key feature of the platform is dashboard management. In short, the solution includes many widgets like gauge, history, and video streaming. Using them, one can generate real-time dynamic dashboards to view data from the IoT system or for reports.
Last but not least, the platform permits analysis starting from predictions made by a machine learning mechanism. With current and historical data, the system can automatically generate auto-machine learning models. The user can then rely on those models in predictions and decision-making.
Our main challenge was to collect vast volumes of data and transform it. To solve this problem, we used a system for mapping the data sources to our models. Thus, we could visually map any data source in our platform’s metrics, events, assets, and custom entities. Besides, we made the platform on the Cassandra database, which is perfect for fast handling vast amounts of data.
Another major challenge came from the customizable dynamic dashboard. We needed it to support dynamic widgets and display the data and diagrams as a technical person from the field required it. To overcome this hurdle, we researched the most popular widgets for our target areas of application. The areas included means building management, production, agriculture, healthcare, and smart cities. Thus, we identified widgets for logs and the graphic components that add value to data visualization and video streaming.
Device management, for controlling the connected devices.
Asset management, for complete systems like an electrical installation or a car.
Dashboard and monitoring for alerts, notifications, and reporting.
Rule engines, to connect the data flows and define actions based on the values of the monitored data.
The platform permits the rapid, no-code development of custom applications.
It can be set in a private cloud, as we have not developed it in a third-party cloud.
First, the platform allows its users to develop connected IoT solutions fast. Thus, they can do everything with widgets and rules without writing a line of code. Second, it brings data protection to the next level. Since we can set the system in a private cloud, you need not share your data unless you want to. Last but not least, the solution can be customized and is scalable and white label. Adding the brand elements of your company and setting the features you need requires almost no effort.
Manufacturing, Agriculture, Smart City, Smart Buildings
Areas of Application
IoT, Big Data
Effort and Duration
3 years, 150 man-months, ongoing