An enterprise IoT implementation checklist

, Category: IT     Twitter Facebook Linkedin Google+  


An enterprise IoT implementation checklist

Is your enterprise IoT solution ready to go? Really? Here are some best practices to keep in mind when adopting or building enterprise IoT products.

IoT technology is expanding its footprint in all major industries as forward-looking companies have already moved from the adoption stage to active implementation of IoT use cases and solutions for their business. IoT innovations are becoming critical for driving operational efficiency on the factory floor and boosting consumer products’ appeal.

IoT-enabled sensors can be found everywhere at oil rigs and large shopping malls, inside brand new cars, on cell towers, and under subway trains. The sheer vastness of IoT-sourced data brings to life an entirely new paradigm of edge computing and analytics. All this makes the IoT an all-pervasive technology phenomena, inextricable from every company’s IT, as well as business, strategy.

Where Is IoT in Corporate Technology Agendas

IoT brings along a number of self-sufficient solutions that offer instant value, such as remote monitoring, predictive maintenance, or location and telemetry tracking. In this regard, certain industries are better positioned to capitalize on IoT capabilities from day one of their implementation. They include industrial, logistics, consumer electronics, healthcare, and agriculture sectors.

Nevertheless, IoT plays even a bigger role as a key enabler of data-driven operation chains, dramatically expanding a pool of valuable data sources and providing the means to effectively collect and process that data. For companies in financial, insurance, retail, or advertising business, IoT opens an unprecedented opportunity to understand customer behavior and create precise business strategies based on gathered insights. In much the same way, equipment manufacturers can tap into usage data and adjust features accordingly or validate warranty claims.

So, on top of its “self-made” solutions, IoT also enables a variety of new cases for Big Data analytics, customer relationship management, supply chain management, and product lifecycle management.

Finally, IoT is essential to every company’s cloud strategy, especially if the company is well-positioned to provide cloud-based services or enable over-the-top applications for its customers, like with Telecoms, but the same relates to consumer electronics and industrial equipment manufacturers. Even hotel chains seize the opportunity to use IoT and cloud-based services to start implementing all-round personalized guest experiences, which are as important as competitive differentiators as well-designed infotainment systems are for modern cars.

All this naturally puts IoT technology at the top of corporate agendas alongside their brand-defining technologies, enterprise back-end systems, data analytics, and cloud applications.

IoT Implementation Challenges

However, to realize IoT's great potential at the enterprise level, companies have to prepare for challenges that they are hardly accustomed to from their previous experiences. These challenges are part and parcel of IoT adoption and should be addressed upfront in a company’s IoT implementation roadmap.

According to numerous market researchers, security remains a top concern among executives that inhibits their IoT ambitions. To effectively minimize the risk for data-sensitive and mission-critical applications running on IoT devices, it is necessary to ensure that a secure private IoT technology, secure deployment, either private or hybrid, and corporate security policies are all in place.

Business Back-End Integration
IoT-enabled devices and sensors allow companies to gather incredible amounts of new data, but many of them fail to create value from that data. If considered separate from key business processes of the company, IoT's potential is severely crippled.

IoT implementations can and should support all money-making operations of the company as well as be extensively used for delivering a personalized customer experience that enhances brand recognition and promotes loyalty. For this purpose, the utilized IoT technology must be built upon extensible architecture that provides straightforward integration with data processing, analytics, and enterprise operations systems. In this case, it will be much easier to bridge new, IoT-enabled sources of data with the actual business development strategy and key applications.

On the other end of the use case are hardware integration and compatibility. A rapidly expanding universe of IoT-primed hardware, connectivity protocols, and purpose-built smart devices pose a significant test to an enterprise's IoT strategy.

In most cases, companies will expect their IoT ecosystem to grow in scope and variety, so the issue of interoperability may become a perennial overhead unless taken into your design from the beginning. This is something that is not too difficult to verify during the IoT technology selection process by judging how open the selected IoT platform is for custom integrations and what type of hardware it already supports.

Technology Evolution
By every forecast on how the IoT revolution is going to develop, what we are observing right now is just the tip of the iceberg. Every attempt to set any strict rules and standards by big commercial technology vendors has failed to take root — standards continue to evolve, and the overall technology landscape is as diverse as it can be.

Since IoT is increasingly becoming a strategic asset for every modern company, the need for its continuous enhancement and adjustment to new business requirements is at the front and center of corporate planning. There are a number of ways to ensure your IoT technology can keep up with the pace of IoT evolution, and it is important to give them a green light right from the start.

IoT implementation Checklist

Considering these major challenges, IoT implementation is not without a twist to a company’s general strategy. Even though it is possible to roll out fast, tactical IoT projects and have your margin, long-term success must be taken into account early in the race to ensure a consistently growing quality and quantity of your IoT solutions.

Define Top-Priority Use Cases
It is wise to start with a clear business strategy on which use cases to implement first and which ones may be long term. A good understanding of the initial scope of required features and capabilities will save you the cost of implementing unnecessary functionality. At the same time, clarified long-term strategy will let you minimize the risk of lock-in with some limited or vendor-specific IoT technology later on. Finally, clearly defined business requirements will let you get through implementation fast and display better ROI for your investors.

Outline Key Technology Assets
Business requirements must be in-depth enough to cover key IoT technology for your initial and long-term IoT initiatives. Among the major technology parts are an IoT platform, hardware, user devices, on-premises and hosted infrastructure, data processing and analytics systems, and business back-end applications. Make sure every component is on the list so that their interoperability and integration is figured out, or at least validated, early in the design.

Clarify Stance on Technology Ownership
Generally speaking, the more strategic IoT is for your business, the more beneficial it is for your company to maintain ownership over its custom IoT technology. Depending on your company’s IoT business strategy defined in the previous stage, at this point, you need to settle on whether you want the freedom and flexibility provided by an owned or open-source IoT technology, or if you are willing to go with a commercial IoT platform as a service and build your business on some vendor’s IoT technology.

It's fair to say that you can change your way later on, but that wouldn’t come without technology transfer costs, extra development cycles, and potentially disrupted workflows while migration takes place.

Plan for Evolution
When your IoT business model has taken shape, it is time to select an IoT platform (in-house, open-source, or commercial) and verify it can go the distance. Make sure your top-priority use cases can be fully supported and that custom functionality can be implemented within a reasonable timeframe and budget.

The selection process may take some time and effort, but chances are the selected technology will become your company’s major business development factor for years to come. Since an IoT platform is a key enabler, it is the first on the list, but hardware, an analytics solution, etc. will need to be decided upon in a similarly rigorous way. You may take into account some extra thoughts on choosing between open-source and commercial IoT platforms.

Ensure End-to-End Integration
To address the interoperability challenge, ensure the selected IoT platform is easily compatible with different hardware, sensors, and gateways, as well as supporting the communication protocols that you are going to need. Ensure all necessary APIs are available and open to use. On the other end, the platform must feature straightforward interfaces to push collected data to enterprise back-end systems and data processing solutions. It's even better if the platform additionally features inbuilt analytics and visualization applications, which can be quickly utilized for your use case.

Address Security and Performance Compliance
Depending on your business environment, comprehensive security and consistent performance may be two major prerequisites for successful IoT implementation. If your company is subject to strict compliance obligations, you need to verify that the selected IoT technology provides modern encryption protection for your data end to end and delivers a consistent level of performance through a high availability mode, elastic scalability, and support for edge deployments.

Consider Private vs. Public IoT Cloud
IoT technology can be accessed from some vendor’s public cloud or deployed at the company’s controlled infrastructure as a private cloud. Both of them have their strong points. However, similar to IoT technology ownership, if IoT is a strategic asset for your enterprise, the benefits of a private IoT deployment will tip the scales to its side. Many companies also use hybrid and multi-cloud approaches to have the best of both types of deployments. The bottom line is that an advanced IoT platform should support deployment flexibility and let you be in control of this process.

Test on Pilot Projects, Take Heed of DevOps
Test your IoT technology on small pilot projects. At this point, also consider cluster setup, provisioning new solutions, user management, etc. Straightforward DevOps operations could speed up upgrade and enhancement of your IoT solutions by many times, so this should be verified during test runs.

Ensure Business Model Continuity
IoT technology is here to stay, and it increasingly permeates companies’ operations, products, and services in every industry. To ensure your enterprise IoT implementation improves at the same steady fashion and brings its entire value to your business, your long-term IoT strategy should be frequently revisited. As to the technology, some types of architectures are just better expandable and customizable than the others. The continuous shift to containers and microservices architectures is a perfect illustration of how the industry is preparing to deal with multitudes of connected devices, systems, and applications.

Outline Potential Partnerships
As the final success factor for IoT implementation, partnerships can speed up your IoT initiatives and multiply their business value. For one thing, partnering with IoT tech companies will allow you to eliminate any gaps in your technology stack. As another advantage, business partnerships will enhance your company’s presence in collateral markets.

IoT platforms comparison: open-source (Kaa IoT) vs. commercial (IoT PaaS) platforms

Technology ownership
IoT PaaS: You rent a bed
Kaa IoT platform: You own the house

Code control
IoT PaaS: Proprietary and opaque
Kaa IoT platform: Open-source and transparent

Cost of ownership
IoT PaaS: Annual fee + usage fee + services + support
Kaa IoT platform: Services + support

IoT PaaS: 3rd party cloud
Kaa IoT platform: Complete freedom: on-premises, cloud, hosted, hybrid

Scalability and security
IoT PaaS: Limited visibility and control
Kaa IoT platform: Full visibility and control

New features
IoT PaaS: Limited control, slow response (reportedly weeks to months)
Kaa IoT platform: Absolute control (open-source), continuous integration, under a week typical response

End-to-end solution
IoT PaaS: Fill in the gaps yourself or engage with a VAR / 3-rd party
Kaa IoT platform: Turnkey, custom solution enablement by KaaIoT

Vendor expertise
IoT PaaS: PaaS
Kaa IoT platform: Full stack - embedded development to data analytics, machine learning, visualization, etc.

Talent attraction
IoT PaaS: Challenging due to the proprietary nature of the platform
Kaa IoT platform: Straightforward due to the open source and reuse of proven and widely adopted open-source components

Professional services
IoT PaaS: Professional services Expensive, shared, on-demand resources
Kaa IoT platform: Dedicated team; proven and cost-efficient offshore methodology

Five key differences

Choosing an IoT platform can be a challenging task in terms of becoming sufficiently knowledgeable about a specific vendor’s offering before getting both feet on board. Nearly as difficult is making an accurate comparison of IoT platforms by the brand since most of the available information is rooted in companies’ unqualifiable marketing hype.

However, two key paradigms for the IoT, which are open-source and commercial platforms, are clearly differentiated on several key lines. Comparing Internet of Things platforms along these lines can help you build a frame of reference for selecting a proper platform for your company’s needs. We believe that unless the factors below are taken under close scrutiny, the selected platform may turn into a waste of time with an unsuitable and costly tool or, in the worst case, into a vendor lock-in.

Kaa is a 100% open-source IoT platform, so it fits perfectly to illustrate what can be the option for IoT enablement other than a commercial PaaS. Kaa also brings its own recipes to address IoT project challenges. While IoT PaaS promise a fast kick-off for IoT application development, Kaa aims to provide a comprehensive IoT platform that guarantees quality and reliability across the entire product lifecycle.

Overall, the following five factors make the biggest difference in our IoT platforms comparison:

Technology ownership
Kaa comes under the Apache 2.0 license, which is a free software license, giving the full ownership over the final solution as well as the solution’s source code to the user. Every bit of everything under the hood is therefore transparent and open to review. It is also easy to modify things or add custom features as soon as the need arises.

IoT PaaS provides a ready-to-go environment for development, which, however, works as a black box and always will. If someday you want to change things around on your own, add a new feature or integrate with some third-party solutions, there will be very few options at your avail in that room you rented. And there is always a risk of being left out in the cold if the PaaS provider’s business goes south.

The math is simple: to use IoT PaaS, you have to pay a rent in the form of annual fee, per-device fee, etc.; Kaa is a free platform. Additional services or support options are available for both.

New features
With Kaa, you have absolute control over the source code of the solution. The most important implication here is that you control the time to market for new features and modules. You can use either in-house development team or request services from KaaIoT, the company behind Kaa - there is no vendor lock-in like with proprietary and PaaS solutions.

With IoT PaaS, you have limited control over new features development, so it might be very important to check that everything fits the bill before subscribing to the service. After that, it’s the PaaS way or the highway.

Interoperability and third-party integration
Building a broad end-to-end solution – where an IoT part is just a link in the chain – is a strategic direction which is not equally feasible with different IoT platforms. IoT PaaS may provide only minimal interfaces for integration with third-party systems and the required integration services may well be out of the PaaS provider’s expertise, which is easy to check beforehand. You will have to fill in the gaps yourself or engage with a third-party development team.

Kaa, as an open-source solution targeted for a broad community of IoT developers, features a variety of established integrations with popular IoT hardware and data processing systems. Kaa is also supported by KaaIoT, a company with in-depth expertise in IoT, embedded development, Big Data, mobility, consumer electronics, and web – allowing you to order a turnkey end-to-end solution at one shop.

The quality of support services is, as a rule, in direct proportion to the amount of provider’s expertise. While most of IoT PaaS providers have been in business for just a few years, KaaIoT is a product IoT development company with over two decades of experience. There’s a large pool of senior software engineers engaged for Kaa development and services, and a cost-efficient offshore methodology is another competitive edge of support services from KaaIoT.