23 Mar 2018
Three-Point-Plan for Choosing the Right IoT Platform
Keith Shea, General Manager, Helix IoT Solutions, Wind River looks at selecting the right platform for use with IoT products
The ubiquity of the Internet can obscure the fact that it is a managed network built on a platform of proven hardware and software. Of course, its openness makes it a target for abuse, but technologies exist to mitigate the impact of attacks, some of which are provided by the platform while others need to be implemented by the user.
Connectivity is synonymous with accessing the services offered by the Internet, which is epitomized by IoT. Choosing a platform to leverage that connectivity to the wider network requires the right strategy.
Without a defined strategy, implementing an IoT platform can quickly become more challenging than necessary. Part of that strategy should define the structural relationship between IoT endpoints, their services and their data, and how that data is managed. For many providers, any IoT strategy will leverage cloud computing and not simply because it provides scalability and flexibility (although it does).
This three-point-plan introduces topics of relevance when choosing the right IoT platform and explains why observing them can reduce your exposure to risk.
1) IoT Lifecycle Management
It should be apparent that the IoT isn’t just about the endpoints, but they are an integral part of an overall objective. Managing those endpoints becomes critical to achieving that objective and so it needs to be addressed in the most effective way possible. Physically, endpoints could be located anywhere, which makes servicing them potentially expensive. Few business models could afford multiple ‘truck-rolls’ to service individual endpoints that are miles apart, because it simply wouldn’t scale as a business.
By putting as much of the endpoint management in the cloud, the need for expensive service visits is minimized. This starts with commissioning the endpoints at installation and provisioning them with the services they need in order to fulfil their own objectives. A connected device is only as good as the platform running it, if it doesn’t provide the capability for commissioning, provisioning and updating remotely it isn’t delivering the full potential of the IoT.
Security is perhaps the primary concern of anyone providing or using IoT services, so provisioning in this context will include establishing secure connections between the cloud and the endpoints. This is rapidly becoming a fundamental requirement of any IoT platform and is integral to the monitoring and management of endpoints. Checking for vulnerabilities in endpoints should be routine, as hackers will target endpoints as weak points in a network’s perimeter.
In order to achieve maximum benefit from IoT services they should be as well integrated into the overall IT framework as possible. This hasn’t always been regarded as important throughout IoT’s evolution, but as the value of data becomes more apparent, operators now understand how the two worlds need to collide. Not all IoT platforms have reached the same level of understanding or integration, which could mask future difficulties when it comes time to decommission endpoints or issue over-the-air upgrades securely.
2) Cloud-centric starting point
Selecting the right IoT platform should form the basis of any IoT strategy; rather than building from the edge in, manufacturers are realising that it starts in the cloud and from there reaches out to endpoints through gateways. Adopting a ‘cloud first’ approach creates a centralized point of control rather than a network of disparate technologies. As there is likely to be a lot of diversity in endpoints it is essential that the centralized management chosen is able to accommodate that diversity; there is little value in creating an IoT platform that is constrained to using suboptimal endpoints simply because they aren’t supported.
This support starts with device drivers; the software components that form the interface between the cloud, gateways and endpoints. A platform that offers support using prebuilt device drivers makes onboarding endpoints simpler and more efficient. As endpoints may be wired or wireless, using Ethernet, Bluetooth or any one of a growing number of protocols, this would include providing compatibility between as many communications protocols as possible.
If endpoints represent the limbs of IoT, the cloud is the brain, so it should be as seamless as possible to allow these component parts to operate together. This kind of out-of-the-box intelligence needs to be inherent to the platform, as too does the ability to allow bidirectional communication between all parts of the platform. Only through this kind of ‘Big Data’ philosophy can it become possible to deliver real-time device management.
Further benefits can be realized when multiple IoT platforms can operate in this way; cloud-to-cloud communications enables an organization’s entire enterprise to become integrated and scalable. This is achieved though robust APIs able to provide a level of abstraction between the differences, and place the focus on the mission of gathering, distributing and acting on data.
By providing APIs the platform allows an organization to maintain a data-centric strategy and avoid the need for one-to-one coding when adding new services, or extended development cycles when expanding the platform to include new endpoints or services. Through support for secure bidirectional communication throughout the platform, data gets to where it needs to be, faster.
3) Network management
By its nature, IoT is a wide area network of disparate and distinct devices, and managing this network involves more than gathering data. It will require a platform that is able to monitor the health of endpoints, diagnose problems and administer remedies.
This level of granularity needs to scale to many thousands of devices in order to support the platform’s lifecycle and can only be achieved if the platform was conceived with the IoT in mind, using two-way communications and remote diagnostics, with over-the-air firmware upgrades. IoT is unlike any network that has gone before it, so the platform needs to reflect that.
This includes the administration of security patches, across all three key elements; the devices, the network and the cloud. In many cases the need to demonstrate compliance with specific security standards has become a requirement.
At the device level this includes supporting chip-level encryption and authentication such as Telecommunications Industry Association (TIA) TR-50 M2M Protocol Standards Series, as well as HTTPS/TLS and secure hash algorithms (TLS 1.2-SHA 256).
Proven methods for network security, such as firewalling and port protection, as well as support for VPNs should also be observed, while at the cloud level the ability to create a secure development and production infrastructure using role- and policy-based access and token-based sessions should be implemented. These are features that should sit alongside a robust and comprehensive auditing philosophy with full record keeping and validation.
The right platform
Developing a platform that integrates wide driver support, a high level of integration and device management with industry-leading security is a challenge that Wind River has addressed through its device management platform. Conceived with meeting the need for providing remote control of IoT devices in a rapidly evolving industry, it provides a secure and scalable architecture for IoT deployment.
Moving to Device Cloud is simplified by its support for zero-touch onboarding of devices. It allows providers to maintain their devices with robust two-way communication, secure field updates and total device management. Its ability to seamlessly integrate with legacy enterprise systems ensures it will complement any organisation’s existing services, while delivering the right IoT platform for future data-centric productivity.
IoT is interpreted in many ways, at its heart it is a network without a hierarchy. While that provides a great deal of freedom it comes with complications. By adopting an IoT platform developed to meet the evolving and specific needs of IoT, businesses are better positioned to overcome the challenges and focus on the opportunities the IoT represents.
Page 1 of 1
About the author
Keith Shea is vice president and general manager of the Helix Cloud IoT solutions business at Wind River, where he is responsible for the company’s SaaS business. His team is chartered with developing and delivering to the market the product portfolio for IoT, as well as managing key customer relationships across broad markets. Keith is also responsible for the ongoing leadership of the Wind River alliance program. Prior to Wind River, he spent more than a decade at Intel Corporation helping the Wind River parent company grow its revenue through strategic software and hardware ventures.
During his tenure with Intel he held assignments in corporate development, product management, marketing, and sales, living and working in both the United States and Europe. Prior to Intel, he worked in the telecommunications software industry with positions in marketing, sales, and product management. He holds a bachelor's degree in economics from Boston College and an MBA from the Wharton School.
Wind River, an Intel company (NASDAQ: INTC), is a world leader in delivering software for the Internet of Things. The company has been pioneering computing inside embedded devices since 1981, and its technology is found in more than 2 billion products. Wind River offers the industry's most comprehensive embedded software portfolio, supported by world-class global professional services and support and a broad partner ecosystem. Wind River delivers the technology and expertise that enables the innovation and deployment of safe, secure, and reliable intelligent systems. To learn more, visit Wind River at www.windriver.com.
Most popular articles in Wireless technology
Share this page
Want more like this? Register for our newsletter