Strengthening its push into the Internet of Things, IBM is making a range of application programming interfaces (APIs) available through its Watson IoT unit and opening up new facilities for the group.
The unit, formed earlier this year with a US$3 billion investment into IoT, will have its global headquarters in Munch, IBM announced Tuesday.
IoT will soon be the largest source of data in the world but, IBM officials point out, almost 90 percent of that information is never acted on -- at least not yet. Many vendors are jumping on the IoT bandwagon and IBM faces a variety of competitors, including tech giants such as Cisco and Microsoft as well nontraditional rivals such as General Electric.
IBM formed the new unit to bring its Watson analytics and cognitive computing services to bear on IoT and highlight its efforts in the technology. The Watson IoT group is courting users and enterprises in the automotive, industry, public utility, insurance, heath care and retail sectors, among others.
"Some of the leading thinking around industrial IoT started in Germany, and there are companies in the automotive industry working on things like connected cars," said Bret Greenstein, vice president of Watson IoT, explaining the decision to put the group's headquarters in Munich.
The Munich facility will bring together 1,000 IBM developers and researchers, and will also serve as an innovation lab for applications involving cognitive computing and IoT, the company said.
The idea behind systems built with the Watson APIs is that, rather than being explicitly programmed, they learn from their interactions with data. The new APIs will be offered as part of IBM Watson IoT Foundation Analytics, the company's IoT cloud platform. They include:
--A Natural Language Processing (NLP) API, designed to let users interact with systems and devices using simple language. Users working on these systems can use human language to query the Watson analytics engine, which could, for example, suggest ways to figure out why a device isn't working as it should;
--A Machine Learning API, which continuously interacts with data input from devices and sensors, "learns" or analyzes it for patterns and prioritizes the data. This API could, for example, analyze weather and traffic along someone's normal route to work and recommend alternatives if necessary;
--A Video and Image Analytics Watson API, which analyzes unstructured data including information from video feeds and photos to detect patterns. The API could be used to figure out the implications or consequences of machine failure in a factory environment;
--The Text Analytics Watson API, which analyzes unstructured textual data such as Twitter feeds, blogs and call center transcripts to, for example, identify appropriate responses to customer feedback.
IBM also announced that it has opened eight new Watson IoT "Client Experience Centers," intended to give clients and business partners access to cognitive computing services and personnel to develop new products. The centers will be in: Beijing; Boeblingen, Germany; Sao Paulo; Seoul; Tokyo; and Massachusetts, North Carolina, and Texas in the U.S.
IBM has scooped up talent, companies and technology in a race to establish a stronghold in IoT and cognitive computing via cloud services. In October, it announced it planned to acquire Weather.com and other digital assets from The Weather Company for the Watson IoT Unit. It has also appointed Harriet Green, former CEO of insurer Thomas Cook, as general manager of the group. In addition, over the past few years in the cloud arena, IBM has acquired a variety of companies including SoftLayer and Cloudant.