2019 IT trend: Enterprises will adopt next-generation platforms to make the most of IoT data

As enterprises transition their physical world to a digital one, smart things become a driving force. Enterprises will implement next-generation platforms in 2019 that analyze large quantities of industry-specific data from the internet of things (IoT). Artificial intelligence (AI) and machine learning (ML) are finding novel correlations between data previously thought to be independent, enabling enterprises to make new discoveries.

This is exciting, as without these platforms, humans would be unable to make these multidimensional correlations — there are too many factors for the human brain to consider. Further, the discoveries from these new correlations can surprise us and conflict with conventional wisdom. We need to shift our focus and learn anew.

For example, precision medicine takes data from new sources (Wifi-connected heart monitors, fitness watches, health records, the human genome) and integrates it with traditional sources (blood chemistry, dietary information). Working with these multiple data inputs will result in more precise diagnoses and treatment plans. Theoretically, someone with low blood pressure, a high red blood cell count and a certain genetic makeup might be more susceptible to BPA toxicity. No longer are diagnoses based on simple one-to-one correlations, but rather on multidimensional correlations.

Autonomous driving is an example of the criticality of real-time IoT data pipelines through analytics and into improved execution. There are an infinite number of random events that can occur when driving from point A to point B. Data must be acquired in real-time and constantly analyzed at the edge (at the car) to execute on an event. The data can also be sent back to the factory (the core platform) for processing with other data to learn about a particular route to supplement the real-time analysis needed for random events. Tesla provides a good example.

Tesla as early adopter

Tesla is a good example of the implementation of a next-generation IoT platform. Tesla has the key advantage of being founded as a digital organization. It looks at the car as if it were a rolling, data-generating device — i.e., computer. Unlocking that data is the key to performance, reliability, operational efficiency, product design and advanced analytics leading to autonomous driving.

Every Tesla manufactured has a digital twin. Data is continuously collected from its 8 onboard cameras, 12 ultrasonic sensors, radar and GPS. This data is sent back to the Gigafactory or to the core platform. Processing and analytics at the core provide intelligence on the workings of the car. Tesla can determine why, when and where it needs to send over the air firmware updates for remote modifications and fixes (performance, reliability and operational efficiency). It can learn from the data and apply the best lessons to new product designs. For example, the Model 3 has a much simpler design, enabling Tesla to build 250,000 units per month relative to the 50,000 per month for the Models S and X.

Processing and intelligence at the edge, onboard the car, provide the promise for autonomous driving.  Autonomous driving requires huge quantities of data, processed in real-time, providing spontaneous and learned responses to real-world scenarios. Neither Tesla, nor any of the other auto makers, can claim to have full autonomy. Tesla is building its own chips both with AMD (an AI chip specifically for autonomous applications) and Nvidia (for high performance computing, analytics and deep learning). Exactly how these chips will be applied is still a mystery. However, autonomy has to be applied at the edge (at the car), which implies a rolling, high-performance, intelligent computer.

Key considerations for enterprises

The rapidly growing computational power of edge devices, along with advancements in connectivity technologies and the need for analytical capabilities both at the core and the edge, will drive adoption of these next-generation IoT platforms. Early adopters must analyze cost versus efficiency, and the scope of implementation. Different industries will require different levels of adoption as well as industry-specific platforms designed to work with the unique data and analytics needs of the industry.

Early adopters must also factor in infrastructure changes; relevance, requirements and advantages of using centralized versus hybrid versus distributed systems; and the readiness of their current environments to migrate or have functionality added. An effective deployment strategy, keeping in mind the business and technological ramifications, will be key to an efficient (and profitable) adoption of next-generation IoT platforms.

Also see the 2019 Digital Trends blog post.

Joan-Carol (JC) Brigham provides in-depth competitive intelligence for strategic deals at DXC Technology. She was an analyst in CSC’s ResearchNetwork for nine years, where she led strategy work and managed much of the launch of industry research in the ResearchNetwork. In addition, she was a principal and business manager analyzing the manufacturing industry. Prior to CSC, Joan-Carol worked at Sun Microsystems in the Services business unit, and at IDC, where she stumbled into market and competitive analysis. @jcbrigham


Mrinal Barman provides competitive intelligence for strategic deals at DXC Technology. He previously worked as an analyst in CSC’s ResearchNetwork, covering the aerospace and defense segment in the manufacturing industry research practice. Mrinal has over five years’ experience as an analyst and has had previous stints with Xchanging and Cognizant. Connect with Mrinal on LinkedIn.

Speak Your Mind


This site uses Akismet to reduce spam. Learn how your comment data is processed.