Tech Insights

Internet of Things: Turn potential into real business value

The Internet of Things (IoT), opens up a world of possibilities for new economic and business value creation. Yet investment by enterprises across industries in IoT also demands the adoption of new approaches to information architecture, application development, and data science.

What is IoT?

IoT refers to the ever-increasing number of physical objects—from small consumer devices to massive industrial equipment—that are now connected to the Internet and are capable of generating data about their operations and the world around them. The number of devices that make up IoT is exploding, with nearly as many devices connected to the Internet today as there are people on the planet. Gartner estimates connected devices in use worldwide will top 20 billion by 2020.

IoT is possible because of a number of converging technology developments. These include the evolution of tiny but powerful data-creating sensors that can be applied to physical objects; near ubiquitous high-speed Internet access to connect devices to the cloud and each other; big data computing power and data science to make sense of IoT data; event-driven architectures to trigger actions in physical, real-world objects; and mobile devices to control and monitor connected devices.

“By the year 2025, $4 to $11 trillion of economic value could be created through the use of the Internet of Things.”

Michael Chui
Partner, McKinsey & Company

Why IoT Matters

IoT investment requires new approaches to information architecture, application development, and data science. Empower your organization with IoT capabilities that help create new, compelling products and services while improving internal efficiencies.

Create and deliver new services to customers

Physical consumer products and devices now connected to the Internet allow enterprises to create and deliver value-add, software-based services that increase customer loyalty, enable cross-sell and upsell opportunities, and reduce customer churn.

Improve operational efficiency and reduce equipment downtime

Data flowing from industrial equipment and other machines enables operators to more easily monitor and manage their operations and even predict equipment failures before they occur so preventative action can be taken.

Intelligently automate operations in real time

Ultimately, IoT and event-driven architectures let enterprises intelligently automate business processes and the operations of connected devices in real time by applying machine learning and predictive analytics to incoming data and triggering next-best actions with no human intervention required.

The Big Differences: IoT Versus Traditional System Data and Architecture
IoT Data and Architecture
Traditional System Data and Architecture
IoT data is distributed. In any given IoT scenario, data is distributed among devices, sometimes thousands of devices, that are often geographically dispersed. Compute and processing power must often take place on network edges. Application data is created and stored centrally. Each application and system creates and stores its own data. Data from disparate applications is then aggregated and physically moved to a central location, such as an enterprise data warehouse, for analysis.
IoT data volume is increasing exponentially. The volume of data in IoT scenarios might start small, but with devices creating data non-stop, 24 hours a day, it doesn’t take long before IoT data volumes become truly massive. This requires a data management and analytics stack that scales to IoT data volumes. Traditional application data growth is linear. Most of the data associated with traditional enterprise applications is created manually. This means the volume of traditional applications data, while growing, is not growing at nearly the pace of machine-generated data. Traditional storage and analytics technologies suffice.
Real-time and predictive analytics are required on IoT data. In order to deliver personalized services to customers and adapt equipment operations to maximize efficiencies, IoT data must be analyzed as it is created and be able to trigger corresponding actions. Rear-view mirror analytics and reporting hinder teams’ ability to be proactive. Traditional applications and systems are often monolithic, containing a web of components that capture data but make it difficult to find and troubleshoot issues in real time. Analytics takes place well after data is created and provides largely backwards-looking views of events and operations.
IoT requires event-driven architecture. An event-driven architecture takes advantage of microservices and allows applications to notify each other of changes in state as they occur and trigger corresponding actions. Traditional architecture is passive. Traditional application architectures are monolithic and don’t support real-time event notification.
IoT usage will change. The one constant in any IoT scenario is change. Data scientists and application developers are always looking for new and innovative IoT use cases, which means they need to continuously adapt existing applications and build new applications using Agile methodologies. Application use is fairly static. Traditional enterprise applications are developed using a waterfall approach to meet existing needs. Applications are rarely updated or changed to meet changing business demands, and new applications take months, sometimes years, to develop and deploy to production.

What to keep in mind if you’re considering IoT

While IoT has potential for enterprises to develop new business models and revenue opportunities, not all enterprises are prepared for the challenges associated with harnessing IoT. Creating, consuming, processing, and analyzing IoT data and operationalizing insights requires making investments in new technologies and related infrastructure, as well as rethinking business models.

Are you prepared for the upfront investment in sensor technology?

Existing devices and equipment (those developed pre-IoT) need to be outfitted with data-generating sensors. The cost and effort required to outfit fleets of objects with sensors is not trivial.

Can your infrastructure handle exploding data volumes?

IoT devices create data—a lot of data. Harnessing it, from storage and processing to analytics and data science, requires a highly scalable, elastic infrastructure and investment in new data technologies.

Can you support an event-driven architecture?

Your infrastructure must not only be scalable and elastic, it must also be event-driven to execute automated actions based on real-time IoT data analysis. IoT-related insights are great, but your infrastructure must support acting on those insights as events occur to create value.

What is your IoT security strategy?

Before embarking on IoT, enterprises must consider and prepare for the data security implications. By definition, IoT data is connected to the Internet, making it susceptible to hackers and subject to regulation, depending on industry. If you don’t know how you will secure IoT data, stop and develop a plan.

Are you prepared to experiment with new business models?

While IoT can enhance existing use cases, its greatest value lies in enabling potential new business models. Enterprises must be prepared to experiment with (and sometimes fail with) new business models and iterate over time.