In this article...

  • From intelligent search to the self-learning supply chain, cognitive systems can transform workplace productivity.
  • Early adopters are using cognitive-enabled search to extend institutional expertise, improve responsiveness and cycle times, and gain more predictive insights.

Regardless of industry, the companies that win in the digital era are those that figure out the shortest path to the best result. That means getting the right information in the right hands at the right time. Those that can do so time and again by unlocking and enabling continual, effective knowledge transfer tend to stay on top.

Those realities are one reason why organizations are turning to cognitive solutions. These technologies combine machine learning, parallel processing and sophisticated analytics to answer questions, provide recommendations and derive predictive insights. Not only can these systems probe vast amounts of data—from text to sound to imagery—they do so at blazing speed, discovering and presenting needed data and information within seconds.

That capability has the potential to transform workplace productivity in the office, at the call center, and out in the field. IBM research shows that, using cognitive capabilities, half of early adopters improve productivity and efficiency, and improve decision making and planning. Nearly half also see cost reductions as a result of cognitive adoption.

Explore Watson Work: Imagine the future you want and redefine everyday work.

Explore Watson Work: Imagine the future you want and redefine everyday work.

Learn more

Making search more intelligent

Companies have been working to bring sleek, relevant and efficient search capabilities to their intranets and portals for years. Most have struggled. Multiple databases, documentation formats, fragmented customer histories, and ballooning product catalogs can make it hard for many corporate search platforms to access and validate needed information. Often, by the time data is captured, cleansed, and tagged, the once-relevant information has staled or become eclipsed by ongoing changes in customer, market or business activity.

One global electric company faced this issue. Poor search quality in their internal and external websites was frustrating customers and employees and contributing to higher-than-average maintenance costs. The company’s existing search platform lacked flexibility. Not only did it have trouble linking data housed in different formats and databases, the platform was unable to read email, PDFs or other types of commonly used software. As Francesc Taxonera, an IBM software client architect, explained, “Plant engineers could lose an hour or more punching in different keywords and search terms to hunt down needed technical documentation. And auditors had to run separate searches in individual PDFs and Excel files to find the data they were looking for.”

To resolve the situation, the company’s head of technology invested in a new, cognitive-enabled search platform capable of interpreting data in myriad formats from across the company’s data repositories. Sixto Vacas, a client technical advisor for IBM, said, “Now auditors can use the search tool to scan everything from text files to email, and engineers will be able to track down needed drawings and specs in commonly used engineering software, such as Enovia and AutoCad.”

That type of intelligent search helps organizations get around the information silos that deliver only a partial view of a given customer or business matter and the manual labor and downtime that result as employees attempt to chase down needed data.

Enhancing customer care

Call centers play a critical role in driving customer satisfaction and loyalty. That makes it all the more concerning that more than half of the 270 billion customer service calls handled annually go unresolved, according to IBM analysis.

With cognitive-enabled tools, customers can pose questions via instant message, text message, email, Web chat, or a dedicated app on their mobile phones, and the system quickly returns answers using informed, evidence-based reasoning. Natural language services, such as an “ask an agent” feature, spare consumers the annoyance of combing through websites or lingering on hold only to be passed through a string of service agents.

At one Asian telecom company, for example, when customers call in with questions about what 4G service or tariff plan is best for them, the cognitive platform can dig out the needed information from a range of sources and deliver the right answers to call center reps within seconds, sharply improving the customer experience. To get the system up and running, call center teams “train” the platform by loading it with product information from databases, catalogs, training manuals, technical support data, terms and conditions, and call center logs, as well as relevant publicly available information. That level of service can create a strong bond, leaving customers feeling more empowered and respected. And because cognitive systems can ingest structured and unstructured data, they’re able to call up information that an agent might miss because the system is looking for semantic links, and not just doing text-matching based on keywords.

Such systems can create profound improvements, but both Taxonera and Vacas caution that it’s important to go into them with a test-and-learn mindset. “Cognitive implementations are complex. They take time to develop, time to train and time to test and improve.” No matter how promising the initiative, there will be speed bumps. Data needs to be hunted down or created from scratch, and the right resources and skills secured. “Build scale by starting small, then refine and grow from there.” says Vacas. “That’s how a business builds believers and takes a new program and new ways of working to scale.”

Build scale by starting small, then refine and grow from there. That’s how a business builds believers and takes a new program and new ways of working to scale.

– Sixto Vacas, Client Technical Advisor, IBM

Build scale by starting small, then refine and grow from there. That’s how a business builds believers and takes a new program and new ways of working to scale.

– Sixto Vacas, Client Technical Advisor, IBM

Managing workflows more efficiently

Workflow bottlenecks can hobble productivity. Within any organization, specialist resources from IT and legal to marketing and customer service can find they spend the greater part of their day responding to queries from other parts of the business. In doing so, they’re fulfilling an important service in extending expertise that other professionals in the business need. But often the steady stream of questions and requests cuts into the time these specialists have to conduct their day job and prevents them from applying their knowledge to create new forms of value.

A leading Spanish bank, for instance, knew that foreign trade was a critically important area for small-and-midsize enterprises, but few branch managers could provide the transaction guidance needed. If a client came into the bank inquiring how to manage export payments to Senegal, for example, the local branch employee had to ring the bank’s central office and speak to a foreign trade specialist. Getting an answer back to the client could take days. Meanwhile, the head of foreign trade was frustrated that his transaction specialists were spending their days answering questions instead of designing new products and services. So the bank created a natural language-enabled cognitive solution that branch employees can access over their desktop or mobile device.

Using this system, employees can ask specific transaction-related questions and receive needed answers in seconds. Alfred Escala, an IBM vice president and client executive, explained. “To create the system, the foreign trade team fed the cognitive platform 2,000 questions that small and midsize clients might pose. About 10% of that information had to be created from scratch—specific decision-making insight that resided only in the minds of the transaction experts. That know-how was the glue in helping the cognitive system make the right connections and judgments, ensuring that the capabilities that were created addressed the most pertinent business issues in the most user-friendly ways.” Now fully operational, the system has dramatically reduced the number of calls flowing into the central office, helped local branch and account teams feel a greater sense of empowerment, and improved customer satisfaction significantly.

Providing early notice of maintenance issues

When a commercial passenger airplane goes out of service, the money begins to leak. There’s the lost revenue from a giant capital asset idling on the tarmac, the unplanned maintenance hours, and the last minute rerouting of flight crews and passengers. Likewise, when there’s an issue on an electrical grid or pipeline, a sprawling port or a plant, identifying the affected area, ordering replacement parts on the fly and placating customers inconvenienced by the problem can tax maintenance staff, strain budgets and lead to inefficient procurement planning.

“Milliseconds are often critical,” says Eduardo Bustamante, COO and CIO of Port of Cartagena, one of the busiest ports in the Americas. Any downtime puts enormous value at stake. The challenge, however, is that with more than 50 industrial cranes moving shipments from more than 14,000 containers, anticipating where a breakdown may occur is very difficult.

To make that problem easier to solve, some companies with remote and hard-to-reach operations are turning to cognitive solutions. These systems, especially when combined with cloud and sensor technologies, provide real-time monitoring and predictive analytics that allow managers to track wear-and-tear and schedule downtime and parts in a far more cost-effective manner.

An aerospace company, for instance, is using the enhanced search and analytics capabilities of its cognitive system to improve supply chain visibility and reduce cycle time, saving millions of dollars on critical parts deliveries. The system lets aircraft technicians search through reams of maintenance records and technical documentation. Now if a worker needs to know what’s causing high hydraulic oil temperatures, the solution identifies historical cases with similar circumstances, finding patterns that point to the root cause of the overheating. The solution now saves the airline manufacturer $36 million a year.

Cognitive solutions can also help companies determine what skills and parts are needed to address maintenance needs. Because these systems are integrated with a company’s supply chain, they are able to check for parts and fast-track orders. The self-learning nature of cognitive systems means they get smarter over time, drawing on an expanding network of information, such as enterprise resource planning (ERP) and customer relationship management (CRM) data, maintenance records, customer calls and more.

At the Port of Cartagena, Bustamente and his team are using their cognitive platform to combine instant and historical views of their operations. The cognitive system they employed allows them to forecast equipment failures and keep ahead of equipment degradation with needed maintenance. “As a container terminal transshipment hub, our port ships goods to almost 600 ports in 136 countries around the world,” said Bustamente. “With cognitive capabilities, we gain immediate insight into the health and operations of our more than 47 rubber tire gantries and 180 trucks.” That helped them cut costs and allows Bustamente’s team to keep vessels and cargo moving smoothly in and out of the port.