The market for embedded systems can look forward to a sustained period of growth thanks to the Internet of Things (IoT). The predicted billions of devices - also known as 'edge nodes' - make up a significant part of the IoT. In some cases, simple devices are critical to unlocking new business models and increasing operational efficiency. Harnessing vast amounts of data is crucial to extract relevant information from hidden data and enable real-time decision making. Every aspect of the IoT relies on reliable connectivity, the provision of cloud services and the constant flow of data from embedded edge nodes.

These embedded devices are often very different. Some are simple sensors to measure temperature or humidity, while other sensor elements are more complex and require more intelligence and system control to make decisions. These more intelligent devices are likely to require an operating system (OS). - More powerful devices even require a real-time operating system (RTOS).

No matter what purpose a particular IoT application needs to fulfill or what architecture it is based on: Manufacturers of such embedded systems have to juggle some business and development considerations to get their solutions to market quickly. Given the openness of the IoT, it is particularly important for networkable devices to get to market as quickly as possible to accelerate market adoption. Products must have features and capabilities that differentiate them from competing products. From a technical perspective, the data security of the devices (device security) is essential. These challenges, combined with the goal of keeping development costs and risks low, put engineering teams to the test.

Operating system for the IoT

By using already certified computing and wireless modules, the hardware challenges can be considerably simplified. The use of an OS/RTOS can also help engineers to develop a data-secure, modular and scalable environment as the basis of their design. The availability of an operating system is even a crucial component in simplifying IoT system development, which can support connectivity and communication between the device and the cloud.

IoT is also responsible for challenges within the traditional development process. The need to quickly develop, deploy and maintain embedded applications has caused many companies to rethink the tools and methods they have been using.
Connecting the development process with the ongoing deployment and update processes, as well as the ability to manage IoT edge devices from remote locations, can bring significant benefits throughout the lifecycle. To enable such a scenario, Windriver offers the Helix Cloud portfolio.

The 'Helix Cloud' elements give engineers access to tools, simulation labs and management platforms that simplify, streamline and automate the way organizations build and operate IoT systems. With Helix Cloud, Wind River delivers a family of Software-as-a-Service (SaaS) products that provide anytime, anywhere access to development tools, labs and deployed devices. This enables developers to develop and manage their IoT technologies throughout the entire lifecycle - from development and testing to implementation and decommissioning.

For example, 'Wind River Helix Device Cloud' is used to administer the devices themselves. Helix App Cloud' provides an environment for testing and developing applications for a specific target platform, while Helix Lab Cloud' provides applications with both hardware, software and simulation.

RTOS for industrial applications

The elements of the Wind River Helix Cloud portfolio.

© Wind River

In addition to providing a cloud-based IDE, the 'Helix App Cloud' offers a close connection to OS/RTOS components that may be needed for a particular design. One example RTOS is 'Rocket' - a free, scalable RTOS developed for 32-bit MCUs. The operating system is suitable for the realization of sensor, industrial and wireless gateway applications. Tailored to devices with limited memory space and low power consumption, 'Rocket' offers functions such as priority-based multithreading, interthread data passing and dynamic memory allocation. Developers can use the operating system to start working quickly - and without having to purchase hardware first. They also have the option of sharing code and project tasks with partners. There is also a developer forum that is accessible within the App Cloud and offers a range of template IoT sensor designs for even faster application development.

With the concept of the Helix Cloud portfolio and already certified computing platforms, IoT applications can be developed in a short space of time. Sensors and actuators can be connected via GPIO/ADC ports depending on the requirements of the design. This technology also makes it possible to establish a platform-based concept for the entire IoT edge device development strategy. Instead of using multiple different board configurations, the ability to base a design on an already proven design ensures that costs and unnecessary effort are minimized over the lifetime of the edge node.

One concept for all IoT devices

For many developers and engineering teams, the ability to establish a platform concept for all IoT devices - whether edge node sensors, gateways or data servers - makes sense from a business perspective. Being able to offer a large product portfolio suitable for everything from small form factor devices and single application to large and more complex systems means choosing an appropriately scalable RTOS. An RTOS that can scale to meet the footprint, functionality and computing power requirements of multiple product classes can help embedded system manufacturers increase the return on their operating system investment, reduce development costs and shorten time-to-market.

The platform concept mentioned above also has another pragmatic reason to be considered - scalability. Of course, IoT topology can be very different. Edge devices can be hundreds or thousands of simple sensors or actuators - from environmental sensors for temperature or humidity to electromechanical or hydraulic actuators. Such devices are very likely to communicate via a gateway or directly with the cloud. The gateway itself may perform local data storage and/or processing before sending data to a cloud-based analytics application.

Many sensors are likely to be battery powered and wirelessly connected. Therefore, they require a corresponding power management capability together with basic network stacks. When considering the challenges associated with scalability - from a simple MCU-based sensor to a multi-core gateway - the benefits of using the same RTOS are revealed as a smart choice.

The age of IoT demands a modular, configurable and extensible RTOS. It complements the features that are already needed today - such as solid real-time performance, low latency and multi-core processor support - with scalability, connectivity, data and functional security (security and safety) and an extended range of functions. Such an RTOS for their future applications gives developers a competitive advantage in the world of IoT, enabling them to bring devices to market faster than before while reducing risks and development and maintenance costs.

Author: Prashant Dubal is Director of Product Management at Wind River.