5 Strategies for Successful IoT Projects

What is your advice to other industry professionals looking to deploy an IoT solution?

There are a few points to think about, which are also presented as an infographic:

1. Do not try to just solve the problem of just connecting and collect data. Go beyond that and ask specific questions about why you are doing this and what business benefit is derived out of it.
2. Avoid implementing a solution for the sake of being an “IoT solution.” Also, question how rapidly the technology can align to the faster changing business needs of the evolving connected world.
3. Avoid vendor lock-in and question standard data models, interoperability, any-to-any protocol handling and most importantly how end-to-end security is handled. Can the platform work with customers using another technology and what is the implementation cost to solve one problem at a time?
4. Move beyond devices and see the scale of digital ecosystems that need to be managed where multi-dimensional relationships become a key challenge.
5. Before spending money on an IoT project, define which stakeholders in the ecosystem benefit today and tomorrow and how that helps drive business growth.

Please describe a recent IoT project you have worked on or have observed that was substantially better, faster, smarter or more efficient than an older technology?

We are in the process of deploying a “smart street light” solution for one of the customers of our strategic partner Tech Mahindra. Ecosystem modeling of the street lights (people who access and manage lights and controllers deployed back-end integration to analytics and business process engines) made a substantial difference in the implementation timeline and the ability to scale the solution to newer cities and regions. The ability to connect streetlights, manage the business metrics and provide business value to the end-customer using ecosystem modeling became a key differentiator when compared to other IoT platform offerings.

What do you see as the biggest potential of the Internet of Things?

Covisint clearly envisions a connected world where it is possible to connect, aggregate, analyze, and share valuable information (which was not possible at scale) across disparate devices, technologies, and frameworks. The concept of a unified data model and removing implementation complexity through an agnostic-platform will define the winning strategy. The potential is huge from a multidimensional perspective that everything humans do to improve quality of life, comfort, and standards will have a direct impact in deriving intelligence and making things smarter. The ability to interact and manage business and life with intelligent systems will become possible with IoT.

What do you see as the biggest problems involving IoT deployments at large?

At this time, the industry is not talking about the need for simulation engines for large scale IoT implementations. The best example is the connected city and how business can rationalize spending without acquiring insight into how the future environment is going to behave – and therefore, what value comes out of it. Combined with the scale and the need for complex relationships (multidimensional) among people, systems, and devices will demand a new paradigm thinking in security and information access modeling. Covisint is focused on solving these two challenges and thus getting recognized as a truly differentiating technology to solve problems in the IoT landscape.

What kind of policy changes or societal shifts do you think are needed for the Internet of Things?

The current state of policies and rules is based on the assumption that humans interact with “less intelligent” devices and systems. As we evolve into a connected world where machine learning and analytics become a key aspect of day-to-day life—with humans relying on machines to do intelligent jobs, the policies need to change. This will be needed to accommodate actions that can be done by machines and how those actions impact human life and the broader society (autonomous driving is an example).