The increasing adoption of IoT has created the next generation of smart devices that are transforming business models and human lives. The proliferation of smart devices has resulted in a complex ecosystem of sensors, software and networks. Customers now need a comprehensive assurance proposition for their IoT Implementations.
A large-scale IoT implementation is comparable to a symphony orchestra, where a group of instrumentalists and their instruments, coordinated by a conductor, produce united pleasurable music. Though the individual instrumentalists are extremely talented, imagine how annoying it would be if one or more of the instruments are out of rhythm. Rehearsal is the key to a successfully coordinated orchestra.
In identical context, though individual elements, such as devices, network connectivity, IoT platforms, cloud infrastructure, data analytics and end-user apps, of an IoT ecosystem are developed and tested separately, the seamless integration and successful coordinated operation of all these elements are necessary to provide a reliable, seamless end-user experience.
In a world of connected everything, IoT devices and platform developers must adopt proven strategies and systems that enable them to deploy reliable IoT technologies faster to market. This interesting technology not only makes the physical world a reality, but also brings a new set of challenges to the community.
IoT-assured best practices
1. Define a scalable IoT strategy. While drafting an IoT implementation plan, enterprises must define both its short-term as well as long-term objectives with focused outcomes. Most often, enterprises are engaged in a small IoT prototype or proof of concept with few critical business cases, and all their assumptions and decisions are made around the use cases and targeted users. Once this system is moved into production and they want to scale, enterprises may hit a roadblock — and it might be too late to realize that the system is not adequately scalable to fulfill an internal product roadmap or network architecture. So, companies end up going back to drawing board to modify their architecture. Some of the decisions, like whether to adopt IoT edge computing or not, should be made during the IoT architecture discussion and everything should be built on top of it. So, it is very critical to define what is expected from an IoT implementation both from a short-term and long-term perspective including the transformation plan.
2. Select the right IoT platform. Needless to say, with the diversified IoT platform market and an array of services from SaaS- to Paas-based IoT and from open source to sophisticated commercial IoT platforms, it is tricky to choose the platform vendor that offers a robust, reliable, feature-rich platform that is cost-efficient. For this reason, it is critical to choose a partner with proven capabilities and trustworthiness. It is important to pay attention to their product roadmap and their ability to provide hotfixes, as well as their success rate in handling production issues, post-production support offerings and maintenance records, and client references from prior implementations from a similar industry. Operating cost estimation has to be drafted jointly based on your implementation plan and their product upgrade strategy to avoid any surprises at a later point of time. Choosing the right platform partner is one of the most important tasks that enterprise IoT architect teams, product management and ops teams should do cohesively.
3. Draft an inclusive fallback plan. Customer IoT experience is like a magic show — once done perfectly, it is appreciated and there is no second time. While building your IoT architecture and cloud components, it is vital to draw all possible situations and draft a bulletproof fallback plan and, more importantly, this plan has to be developed, tested and approved before the entire IoT system is pushed to production. Integrated fallback plans include things like cloud server backup, alternative network switches, and IoT edge and local computing facility for mission-critical applications such as healthcare or nuclear power plants. In nutshell, have a fallback plan that is robust and reliable in case of an emergency.
4. Choose quality over features. When it comes to keeping the pace with platform vendors, it is good to stick to the basics and choose quality over features. Look for stable features in the IoT platform and build your system on top of it. But take caution when accepting new build/patch releases and do not rush for rich features as IoT platforms are maturing in parallel — it is important to give some time before choosing the right version of the platform for your needs.
5. Embed quality across the software development lifecycle. Implementing quality gates at each layer and different phases of IoT implementation is a critical success factor. Three core elements for IoT quality are:
- Test infrastructure — A production-like test infrastructure that can support not only preproduction testing of devices and platforms, but is also suitable for post-production situations is important. This continuous pipeline should support periodic interoperability and integration testing from device and sensor OEMs, as well as patch upgrade requests from IoT platform vendors without disturbing ongoing product enhancement. This is a living environment that needs to scale based on your production.
- Custom tools and techniques — Every IoT implementation has unique devices, networks, platform architectures and tools that can mimic IoT devices and platform behaviors, like digital twins. These must be tested for extreme real-life situations before they are ready to deploy into production.
- Integration test strategy — After the successful testing of individual components, executing end-to-end integration testing, without any compromises, is a must. Automated periodic integration testing is a best practice that can unearth defects earlier and help accelerate production deployment.
As IoT platforms and device ecosystems have matured during recent years, the acceptance of more and more IoT use cases and products has gained momentum across industries. As the market evolves, new technologies, protocols, platforms, ecosystems and processes for building IoT use cases has also evolved to a great extent. Nevertheless, being successful in IoT is defined as a balancing act between the stability of the system and the cost of operation — and, yes, keeping IoT dreams alive still remain a key challenge due to various reasons. While most of the technical challenges are being addressed every day, successful IoT implementations require a thorough project management plan blended with industry best practices.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.