Watson is the answer to the exponential growth in complexity of system development

The Internet of Things (IoT) promises to deliver a tremendous amount of data from the billions of sensors that will serve as the eyes and ears of the IoT. The number of sensors is already growing dramatically, and it is happening in every business segment. In the oil and gas industry, a modern facility has more than 80,000 sensors and a single reservoir will produce more than 15 Petabytes (PB) of data over its lifetime. Close to 700 million smart meters will be installed in the energy and utilities sector by 2017, and these meters will be producing more than 280 PB of new data that needs to be sorted, categorized, analyzed, and acted upon.

By 2020, more than 50% of the appliances on the market will be connected — a dramatic growth from the 1% connected today. Also, by 2020, over 75% of the world’s automobiles will be “connected” and will produce 350 Megabytes (MB) of data per second. Some of this data will be critical to passenger safety, but where in the 350 MB of data per second will that data be found?

Fig. 1: Reacting to the dramatic growth in IoT data will challenge traditional business models.  (Image Source: IBM)

The IoT will provide an unimaginable avalanche of data for businesses to sort, categorize, and act upon. Traditionally, this information is used to create a competitive advantage. The business with the most information wins. Unfortunately, massive amounts of IoT data can also bury a traditional business when it can’t find the important, actionable information hidden within their IoT data. The new competitive advantage now requires not only massive amounts of data, but the processes and techniques to extract critical information and then identify the associated insights that can be turned into competitive advantages.

In addition to the growth in IoT data, there is an even more concerning growth in system complexity. With the massive interconnectivity that the IoT provides and the associated device interactions that result, system complexity grows at an exponential rate. As just one example of this growth in complexity, consider the difficulty in understanding all of the unintended consequences that complex device interactions can generate. Users may ask devices to do things in unexpected ways that will impact other connected devices and, thus, the entire system. Bandwidth requirements may shift, response times may extend, and the user experience may degrade. This can result in users trying unexpected ”work-arounds” that further impact system performance.

Increasing complexity impacts product development

Providing security, reliability, and quality of service while adapting quickly to changing user requirements in this new, more complex environment is overwhelming previous product development tools and methodologies. The traditional development approach has used a linear process, with requirements defined up front and various development stages used to create an implementation that balances all the requirements and meets various cost, form factor, and compliance regulations without slipping the schedule. New requirements invariably show up late in the development process and adversely impact schedules, costs, and quality. With the exponential increase in complexity that IoT systems create, we need a dramatically different approach to product development.

Instead of a fixed, linear methodology for product and system development, what we need is a continuous engineering methodology, as illustrated in Fig. 2 below. This approach assimilates all aspects of product and system engineering into an integrated whole, with requirements driving the design, verification, implementation, testing, manufacturing, and deployment activities continuously, not just at the beginning of the project. Additionally, once IoT systems are deployed, new data is generated that can be used to monitor performance and provide maintenance proactively. The data can also be checked to determine if user experiences and operations are generating expected outcomes. If new use cases and strategies emerge from the analysis, new requirements can be provided at any point in the development process to accelerate system improvements and deliver increased customer satisfaction. This approach delivers a dramatically different system to the user and creates a compelling competitive advantage in the marketplace.

Fig. 2: Continuous engineering redesigns the traditional development process. (Image Source: IBM)

IBM’s Watson IoT Platform drives continuous engineering methodologies

Creating a successful continuous engineering methodology hinges on the availability of actionable information during the development process. This data can come from the mountain of documents produced in previous projects, including specifications, requirements, test results, and simulation — just to mention a few — as well as from data received during operation (the top part of the process shown in Fig. 2) that captures operational metrics for wear, user interactions, error conditions, requirement shifts, and many other associated system measurements. Categorizing, sorting, prioritizing, and analyzing these mountains of data would be an overwhelming effort without significant help.

IBM’s Watson IoT Platform provides a set of capabilities that directly address the complexity challenges that the IoT brings to engineering development. Watson isn’t explicitly programmed. It learns from experiences with the environment and interactions with people. It provides a cognitive approach to systems and processes and understands goals and objectives. It can integrate and analyze mountains of data and can find the important connections. For engineering development, Watson might act as the “guy a couple of cubicles over” who was on every previous project and knows where all of the important and relevant documents, test procedures, and simulations are located. With Watson, this “institutional memory” can be extracted directly from development data archives and used as needed. Watson never goes on vacation, never quits, and never retires.

Conclusion

The IoT presents many opportunities and challenges to traditional business units as well as engineering development methodologies — most notably, the increase in data availability. The sudden proliferation of information that IoT offers must be addressed accordingly — on both fronts — in order to ensure long-term sustainability. Companies will soon be expected to evolve their traditional approaches and cognitive computing capabilities, like those available from IBM’s Watson IoT Platform, in order to better sift through the avalanche of available data and identify the key insights that can be turned into market-winning competitive advantages.

Next steps and additional information:


Related


Making things come alive in a secure way

Considering the security of one single business application is not enough anymore. With the IoT, multiple applications across multiple industries can share and exchange data across different types...

Monitoring and Securing the Smart Grid

Smart grids use phasor measurement units (PMUs) to assess and manage independently operated grid systems. National Instruments, Hewlett Packard Enterprise, and OSIsoft created a "best of bree...

IoT Kit Pairs Gateway with Cloud Platform

Few companies have expertise in both embedded design and Big Data—or know how to combine these domains. The Advantech IoT Gateway Starter Kit simplifies matters. Find out how this gateway co...

Monetizing the IoT: Show Me the Money

The Internet of Things (IoT) is disrupting entire industries around the world. There is a lot of ‘new’ –including new devices, software, services – along with new opportunities...

Vending Machine IoT Makeovers

Operators of large vending machine fleets need ways to update these machines to profit from Internet of Things advantages. Explore a retrofit solution engineered by Bsquare and AAEON that uses a I...

SAS Analytics for the Internet of Things

With IoT device data on the rise, SAS Analytics has the software solution. Designed to serve a variety of industries, SAS Analytics for IoT supports analytics at the edge – as well as analys...

Protecting IP in the IoT with Trusted Computing

In this on-demand webinar Steve Hanna, Senior Principal at Infineon Technologies, and Guenther Fischer, Senior Consultant Licensing and Protection at Wibu-Systems, present a broad picture that hig...

LTE-A for Data Intensive IoT Applications

LTE-Advanced (LTE-A) is the latest enhancement to LTE. As its name implies, LTE-A is well suited to applications that need the next level of performance. Find out how to improve practical data rates a...

The Rise of the IoT Analytics Gateway

Collecting large amounts of edge data and sending it to the cloud can slow network performance and increase storage costs. Learn how equipping IoT gateways with dual-core processors enables analyz...

Creating embedded systems enhanced with IoT

Engineering a bridge between the physical world and the online universe Engineers have gained some design flexibility with the advent of the Internet of Things (IoT) and the many options for produc...

Turning Healthcare Data into Value

Everyone wants to reduce healthcare costs and improve care. Tapping Big Data helps accomplish both. Discover how an IoT health analytics platform monitors patient populations and helps improve wel...

Six Hidden Costs in a 99 Cent Wireless SoC

When adding wireless functionality, the two approaches generally available include using a system-on-a-chip (SoC) or a module. The SoC is lower cost, but comes with various unknowns of how to desi...

What’s the right security for IoT?

As our world transforms into a highly connected architecture, security has become a strategic concern. Impacts extend beyond financial losses into the physical realm. This video will give you a quick ...

Licensing and Security for the Internet of Things

The IoT needs much stronger protection than the closed systems of the past. Safeguards against reverse engineering and manipulation are required along the entire design chain – from the hard...

M2M Communications and the Internet-of-Things

Machine-to-Machine (M2M) Communications used to be a niche topic. But today it is regarded as the key technology of a networked, mobile future. On this microsite we present a collection of White Paper...

Planning a scalable long-term wireless strategy

For companies with global deployments, there is no single right answer for planning a scalable long-term wireless strategy. The best option, focus on solutions that allow for reuse of development, tes...

IoT makes Smart Foundation for vacation resort

Vacation resorts are looking for ways to help conserve water and energy while improving guest comfort and enjoyment. With the rapid growth of the Internet of Things and the increasing availability...

Connect building systems to the IoT

In this article, we will explain how smart buildings can connect to the cloud, and show how building owners, equipment providers, and service companies can benefit from this connectivity. Specific...

IoT takes mass transit to a new level

The Internet of Things is critical to Bus Rapid Transit (BTR). It enables real-time collection and transmission of data, helping improve fleet management, scheduling, ticketing, safety, and advert...


Embedded Systems Safety & Security Survey

Andrew Girson, CEO of Barr Group, shares key findings of their 2017 Embedded Systems Safety & Security Survey. With responses from more than 1,700 qualified engineers worldwide, the independen...

What you say is what you get: an Ada story

Over the past 30+ years, Ada technology has matured into a unique toolset allowing programmers to achieve software reliability at a very affordable cost. It’s available for small microcontro...

Addressing IoT impact on software engineering

Manufacturers need to carefully evaluate the cyber threats and the level of exposure of IoT devices. New levels of software integrity can only be achieved if teams can eliminate both accidental co...