Until now, the analytics world has mainly viewed the internet of things as an opportunity for gathering even more data, which can be used in predictive models and other analytics initiatives. But it turns out advanced analytics and cognitive computing systems can also play a role in making devices smarter.
At IBM's World of Watson event in Las Vegas, office supply company Staples introduced a new customer service tool that leverages Watson's natural language processing (NLP) capabilities and text analytics.
The device is built to resemble the Easy Button in the Framingham, Mass., company's marketing campaigns. Pressing the button activates a microphone. Users can then use voice commands to place orders or check the status of deliveries. These features can also be accessed through more typical platforms, such as mobile devices and web browsers.
Faisal Masud, Staples' executive vice president of global e-commerce and chief digital officer, said he sees these kinds of cognitive-based customer service and ordering systems as a key part of 21st century business.
"We're seeing applications across the board," he said. "We're seeing applications across customer service, but also across e-commerce, where it could be the operating system."
Analyst firm Gartner forecast 6.4 billion connected devices will be used worldwide in 2016, jumping to 20.8 billion by 2020. But many of those devices will be relatively dumb, including things like fitness trackers, connected speakers and cameras. The real value from internet of things (IoT) devices could come from backing up a data connection with cognitive computing systems. This turns a connected device from something that strictly generates data into something much more interactive.
Thomas Jefferson University Hospitals is using a similar approach to help improve the patient experience. The health system recently unveiled a speaker and microphone system that can be embedded in patient rooms. Patients can speak questions about hospital services or commands that control the room's heating and cooling, lighting and entertainment systems.
The system is built around Watson's NLP and classifier engines. Together, these tools decipher patients' requests and then, through tie-ins to the building's HVAC and entertainment systems, take an action. The health system is currently training the machine learning algorithms that process voice commands and will begin rolling it out to patient rooms once it achieves accuracy in processing requests of over 90%.
Speaking at the conference, Neil Gomes, the health system's vice president for technology innovation and consumer experience, said he believes this project is laying a foundation for improvements in the delivery of care.
"Cognitive and IoT, that's where we have to go to," he said.
General Motors announced a new link between its OnStar vehicle connectivity system and the IBM Watson cognitive computing system. The new platform lets users set up reminders for tasks to complete on a certain trip, receive traffic alerts while on their way to work or pay for gas from the dashboard system. These services tie into Watson on the back end and learn drivers' habits to deliver more personal recommendations and services over time.
GM's CEO Mary Barra said the combination of IoT and cognitive computing systems will lead to greater changes than the auto industry has ever seen, and it's important for the company to keep pace.
"I believe the next five years will bring more changes than the last 50," she said. "Now, we can make cars a cognitive platform that learns who you are."
Cognitive computing may be link between big data and analytics
Still no straightforward answers for how businesses will use IBM cognitive computing system
IBM still working to deliver on cognitive computing hype