How the cloud can improve embedded systems
Embedded systems are isolated and kept independent from one another, but the cloud can help bring these systems together to give manufacturers greater visability and operational capabilities.
- Understand the benefits of connecting embedded systems to the cloud.
- Learn about the potential risks and challenges that come with connecting these devices.
- Explore why hardware design and communication protocol selection are important.
Embedded systems insights
- Tying embedded systems to the cloud helps manufacturers better understand how operations are performing and where they can improve.
- Storing the data securely is a concern because many of these devices are older and weren’t designed to be cybersecure, which means manufacturers have to take additional precautions when gathering and storing the data.
Not too long ago, factory floors were swamped with machines equipped with purpose-built software that used closed architecture unique to each device. These were isolated devices not connected to any cloud network and considered secure, safe, reliable and efficient for their intended designs.
However, the proliferation of the Internet of Things (IoT) and cloud advances, the growing need for intelligent sensing and control solutions and the promised speed of emerging 5G wireless infrastructure are prompting manufacturers to rethink its industrial landscape. Manufacturers recognize the value analytics from the cloud brings in optimizing and introducing new levels of efficiencies that are difficult to replicate with traditional disconnected embedded devices.
How embedded devices bring value to industrial operations
The design of legacy embedded systems was hardware confined. This means the data remained at the end of the device or connected to other select embedded systems devices. While this made the systems more secure from the vulnerability of the cloud and got the job done efficiently, they also reduced the data’s lifecycle.
Giving way to a software-defined open architecture connected to the cloud enables resource sharing across geographies, time zones, users or even applications. This also allows easy storage, maintenance and updates in the future. Consider a connected pacemaker, which is embedded with sensors that gather vital patient data such as the heart rate, pulse rate and reading from implants. This data is then stored in the cloud, ensuring it is not constrained to any device but available to doctors for review on their wireless devices or even stored for future reference.
In an industrial setup where multiple propriety embedded systems run parallel, there could be a system for telematics, one for braking and control, one for radar, one for connectivity and many others. Each has its operating system (OS), dedicated silicon and certification process. These isolated systems may not be connected, but they are located in the same cloud, which enables them to work together and share information with the manufacturer or the machinery. They are stored in the same cloud without additional cost. With distributed systems in a more significant site, access to all the systems for maintenance and updates gets simpler. This shift leads to reduced costs and faster time to market.
Three ways to improve embedded system integration to the cloud
Theoretically, these benefits from a connected embedded system are compelling enough for manufacturers to embrace this transformation, far outweighing the concerns of security and reliability that isolated embedded systems deliver. However, many underlying challenges could impact the health of the whole unit.
1. Governance and cybersecurity.
Security is one of the paramount concerns for even a standalone system connected to the cloud. Imagine how the intricacy exponentiates for a highly interconnected system. Being connected to the cloud and with one another makes embedded systems much more vulnerable to security threats than siloed systems. Disastrous data breach cases in the recent past demonstrate how hackers gain unauthorized access from one system to another.
Other commonly used security protocols for managing embedded systems as well as network equipment are secure sockets layer (SSL) and transport layer security (TLS), simple network management protocol (SNMP) and more. These protocols provide mechanisms for secure secret key exchange and a layer of security via remote configuration, view-based access control, and logical contexts for administration and authentication.
Enterprise solutions are also widely available to connect embedded devices to the cloud securely. In addition, there is a rising trend to shift from system-level certification to component level to ensure the safety of embedded systems.
2. Designing hardware properly for embedded systems.
Hardware design is a critical component that needs to be considered for embedded systems. An inflexible design, for example, hinders multiple applications running over embedded systems or creates difficulty in adapting to a new environment. Again, inadequate functional safety in design can make task scheduling and timing difficult. Hardware design also impacts power consumption, cost and time to market, which diminishes embedded systems’ performance while connecting to the cloud.
Designing hardware for embedded services requires careful planning keeping all these challenges in mind. This planning involves hardware platform design, firmware/embedded software, testing, wireless design and platform software design.
These systems need to be flexible to ensure smooth integration in new environments and with new services. The designer must build a robust and reliable system with cryptographic algorithms and security procedures. An efficient system architecture design should limit power consumption and enable the embedded device to work efficiently even with low power. For reliable product designs, conducting in-depth embedded hardware testing, verification and validations is necessary.
3. Choose the right communication protocol.
Loss of communication among different sensors and actuators is a common problem while connecting embedded systems to the cloud. The problem is exacerbated when the control system needs to use multiple real-time network communication protocols that do not allow direct routing over a network. The data needs to be embedded in routable packages via a gateway without lowering the performance.
Middleware also can be deployed in a cloud infrastructure for interfacing with embedded systems and allowing remote firmware updates in these embedded systems. There are communication methods via gateways. Communication via smartphones and tablets also has become common.
While some control loops have been moved to cloud services, there is still a lot to explore to uncover the full potential of connecting embedded systems to the cloud. This transition cannot be a simple extension of today’s processes. It is a paradigm shift requiring a complete reimagination of how manufacturing and production systems are designed to simplify and effectively develop cloud-based embedded development.
Sacheen Patil, vice president and global head – IoT and embedded practice center of excellence (CoE), Yash Technologies. Edited by Chris Vavra, web content manager, Control Engineering, CFE Media and Technology, firstname.lastname@example.org.
Keywords: embedded systems, cloud technologies
See additional edge and cloud computing stories at https://www.controleng.com/edge-cloud-computing/
How is cloud computing changing the way you integrate devices, systems and manufacturing analytics?