Cognitive computing: Hello Watson on the shop floor

Share this post:

Cognitive systems are fundamentally different

Cognitive computing refers to systems that learn at scale, reason with purpose and interact with humans naturally. Rather than being explicitly programmed, they learn and reason from their interactions with us and from their experiences with their environment. They are made possible by advances in a number of scientific fields over the past half-century, and are different in important ways from the information systems that preceded them.

From tabulation via programmable to cognitive: the three eras

IBM has been working on the foundations of cognitive computing technology for decades, combining more than a dozen disciplines of advanced computer science with 100 years of business expertise.

Figure 1: Three eras of computing

Figure 1: Three eras of computing

The Tabulating Era (1900s — 1940s)

The birth of computing consisted of single-purpose mechanical systems that counted, using punched cards to input and store data, and to eventually instruct the machine what to do (albeit in a primitive way). Tabulation machines were essentially calculators that supported the scaling of both business and society, helping to organize, understand, and manage everything from population growth to the advancement of a global economy

The Programming Era (1950s — onward)

The shift from mechanical tabulators to electronic systems began during World War II, driven by military and scientific needs. Following the war, digital “computers” evolved rapidly and moved into businesses and governments. They performed if/then logical operations and loops, with instructions coded in software. Originally built around vacuum tubes, they were given a huge boost by the invention of the transistor and the microprocessor, which came to demonstrate “Moore’s Law,” doubling in capacity and speed every 18 months for six decades. Everything we now know as a computing device—from the mainframe to the personal computer, to the smartphone and tablet — is a programmable computer.

The Cognitive Era (since 2011)

The Cognitive Era is the next step in the application of science to understand nature and improve the human condition. Within the scientific community—as opposed to the media and popular entertainment —the verdict is in. There is broad agreement on the importance of pursuing a cognitive future, along with recognition of the need to develop the technology responsibly. The potential for something beyond programmable systems was foreseen as far back as 1960, when computing pioneer J.C.R. Licklider wrote his seminal paper “Man-Computer Symbiosis.”

Data and information explosion – 1 ‘brontobyte’ of data in 2020 – amount of data to be managed

Gartner estimates the world’s information will grow by 800 percent in the next five years – with 80 percent of that data being unstructured. Unstructured data comes from texts, photos, videos, books and manuals. It is data hidden in aromas, tastes, textures and vibrations. It comes from our own activities, and from a planet being pervasively instrumented.

Figure 2: Unstructured data — “dark data” — accounts for 80% of all data generated today.

Figure 2: Unstructured data — “dark data” — accounts for 80% of all data generated today.

In a global economy and society where value increasingly comes from information, knowledge and services, this data represents the most abundant, valuable and complex raw material in the world. And until now, we have not had the means to mine it effectively.

New ways of managing the data to gain knowledge

Programmable systems are based on rules that shepherd data through a series of predetermined processes to arrive at outcomes. While they are powerful and complex, they are deterministic—thriving on structured data, but incapable of processing qualitative or unpredictable input. This rigidity limits their usefulness in addressing many aspects of a complex, emergent world, where ambiguity and uncertainty abound.

Cognitive systems are probabilistic, meaning they are designed to adapt and make sense of the complexity and unpredictability of unstructured information. They can “read” text, “see” images and “hear” natural speech. And they interpret that information, organize it and offer explanations of what it means, along with the rationale for their conclusions. They do not offer definitive answers. In fact, they do not “know” the answer. Rather, they are designed to weigh information and ideas from multiple sources, to reason, and then offer hypotheses for consideration. A cognitive system assigns a confidence level to each potential insight or answer.

Realizing the true potential of cognitive

Now we are seeing firsthand the impact of cognitive computing – its ability to transform businesses, governments and society. The true potential for the cognitive era can be realized by combining the data analytics and statistical reasoning of machines with uniquely human qualities, such as self-directed goals, common sense and ethical values.

This is what Watson is doing now – helping banks to analyze customer requests and financial data to surface insights to help them make investment recommendations. Companies in heavily regulated industries are querying the system to keep up with ever-changing legislation and standards of compliance. In the health industry, oncologists are testing ways in which cognitive systems can help interpret cancer patients’ clinical information and identify individualized, evidence-based treatment options that leverage specialists’ experience and research.

Industry 4.0: Embracing cognitive IoT for manufacturing

Improving operations and increasing competitive differentiation are top of mind among manufacturing organizations. By utilizing the power of cognitive capabilities, IoT for manufacturing can help harness and mine the influx of information—making shop floors more cognitive through effective processing, analysis and operational optimization.

Industry 4.0 offers organizations new ways to adapt or improve processes for better quality or to meet market demand. Industry 4.0 goes beyond automation into instrumentation of the shop floor via IoT-aware devices, such as sensors, beacons or RFID. The use of these advancements helps to achieve a granular level of integration – where data and information are delivered in real-time.

Here are a few examples to illustrate how interoperability between machines and workers can be enhanced; and, where higher levels of quality and productivity can be achieved.

Integrating machines, humans, manufacturing processes and back-end systems

Increased customization is a major challenge for manufacturing. The adoption of the digital factory concept is how John Deere is working to achieve this. John Deere is adapting IoT, analytics and cognitive to their own needs through their Smart Manufacturing Platform (SMP) which spans three layers – the edge, plant and cloud. For the organization, the main goal of the SMP is seamless integration of machines, humans, manufacturing processes and back-end systems.

On the shop floor in each factory, by combining lean manufacturing principles – implemented over decades, with current Industry 4.0 technology, they have achieved the realization of three main use cases:

  • Digital Build Paper as the implementation of a paperless factory for enhanced flexibility in order re-sequencing, prioritization and identification
  • Worker Assistance with focus on intelligent worker and service technician support for error prevention and decreasing costs, time and effort for rework
  • Sense & Act addressing productivity increase by avoiding non-value-adding activities like scanning operations by leveraging RFID-based localization technology.

The process begins with order picking, where the necessary parts are shown to staff through an ERP (Enterprise Resources Planning) system. After the worker selects the part, the withdrawal is confirmed and then reported to the central platform in real time. To ensure the required materials are available at all times, and IT system also produces documentation automatically – moving manufacturing organization one step closer to being a paperless factory.  Watch the video to see the Smart Manufacturing Platform in action.

Manufacturing organizations like John Deere are also tapping into Watson’s cognitive capabilities, by integrating cognitive maintenance into their processes. For example, using image recognition capabilities, an operator can take a photograph of a workstation and Watson’s image recognition algorithm will help to determine what’s wrong with machinery, while also providing corrective actions to fix the fault. Operators can also use Watson speech recognition capabilities for hands-free voice commands.

Visual Inspection

Visual inspection is another use case example. Cognitive Visual inspection helps organizations to detect defects via real-time production images that are captured through an ABB solution, and then analyzed using IBM Watson IoT for Manufacturing. Previously, these inspections were done semi-manually, which was often a slow and error-prone process.

Figure 3: Cognitive visual inspection

Figure 3: Cognitive visual inspection

Adaptive robotic maintenance

In another example, we can see how cognitive is applied to equipment maintenance to enable different interactive services to assist machine technicians and operators. By directly integrating edge devices – such as machines and robots – in the cloud using Watson IoT Platform, manufacturers can develop personalized products and services, improve operations, reduce costs and avoid the risk of downtime. By accessing different Watson services, in addition to other APIs on IBM Bluemix, a technician or operator is able to take advantage of analytic functions, predictive maintenance, and visualization of information in a dashboard.

Figure 4: Cognitive robot maintenance example

Figure 4: Cognitive robot maintenance example


Realize the potential of cognitive computing by taking logical steps

Figure 5: Embracing Cognitive IoT for manufacturing relies on logical steps to value.

Figure 5: Embracing Cognitive IoT for manufacturing relies on logical steps to value.

Start by gathering the data – data can come from systems such as EAM, ERP, MES. It can also come straight from equipment, robots and sensors – by instrumenting equipment and assets to collect data, and, organize and include existing or historical data from previous years. Assets needs to be connected, outfitted with sensor for data to be gathered.

Next, visualize the patterns. This can be done through dashboards, equipment user interfaces (UI), and other representations. Seeing the data through common dashboards can make a huge difference in time required to carry out repairs. Watson IoT solutions can be used to quickly build dashboards for data visualization.

Advance to analytics. Advanced analytics help organizations to gain insights from data and information – to produce models and make predictive recommendations. This data can be enriched with the addition of data from other sources – systems, sensors, external conditions and environmental data such as weather. Organizations can use analytical models to predict equipment failures and provide recommendations.

The fourth stage is cognitive fusion – where models can be refined with machine learning, and other cognitive functions and capabilities can be applied to improve engagement – for example the use of speech, video, image to diagnose complex problems.

Seeing is believing – check out our 30 day trial offers.

  • 30 day trial of IBM Prescriptive Quality for Manufacturing. When it comes to cost of quality, there’s always room for improvement. Prescriptive Quality for Manufacturing provides earlier, more definitive identification of quality problems in comparison to traditional statistical process control methods. If you are interested in employing prescriptive analytics to improve quality of manufacturing processes, materials, components and products, register for a 30 day free trial.
  • 30 day trial of IBM Prescriptive Maintenance for Manufacturing. Manage the asset reliability risks and reduce your plant operations exposure. Using predictive models which determine the likelihood of asset failure and time to failure based on statistical modeling can help to identify top failure modes; recommend optimal maintenance schedules and procedures; help with ongoing optimization through machine learning and physics based models. If you are interested in improving process and productivity quality, register for a 30 day trial. It’s free.

Learn more:

Learn more about IBM Watson IoT solutions for Manufacturing.

End notes

Kelly, John E, Computing, cognition and the future of knowing: How humans and machines are forging a new age of understanding, IBM, 2015

Add Comment
No Comments

Leave a Reply

Your email address will not be published.Required fields are marked *

More Manufacturing Stories
By Dan Bigos on January 24, 2017

The Industrial Internet of Things is full of transformational potential

Plant Engineering a “goto” source for practitioners to maintain currency on trends, best practices, new technologies, and operations perspectives grounded in reality devotes a healthy portion of content to the topic of Industrial Internet of Things – IIOT. And in the manufacturing world, IIoT portends tremendous transformation. The value of the IIoT The McKinsey report […]

Continue reading

By Pete Karns on July 26, 2016

Understanding asset health with Maximo

Change and transformation in business is inevitable. Last December I spoke at the International Maintenance Conference about how advancements in technology have disrupted industries over time – and why it is critical to utilize advancements for improvements, or get left behind. With all the hype around the Internet of Things – we need to look […]

Continue reading

By Dan Bigos on April 20, 2017

Predictive maintenance is just one of IBM’s IoT for Manufacturing strengths

Many companies make a claim to the predictive maintenance market. Many more than you might imagine. In fact, IoT Analytics in its Predictive Maintenance Market Report 2017-22 has identified at least 110 technology firms that have worked on predictive maintenance implementations. These firms address various parts of the overall predictive maintenance solution – condition monitoring […]

Continue reading