While the internet of things (IoT) headlines obsess over wearables, the really big money will be found in far less glittering platforms being developed behind the scenes by major powerhouses like IBM and General Electric. While operators catch their customers’ attention with connected cars and smart home hubs, companies like Verizon and AT&T know that they could achieve far greater business model transformation by partnering with these ‘industrial internet’ initiatives.
Movements like the GE-led Industrial Internet Consortium (IIC) revolve around connecting all kinds of vertical market and industrial ‘things’ – from aeroplanes to warehouse pallets to components – to one another, to the internet and to big data systems in the cloud, using a variety of sensors, tags and short- or long-range wireless connections. These systems can be turned into gold by up-ending supply chain arrangements, making them more responsive and efficient; by tracking and maintaining even the smallest asset accurately; and of course by using all that data to improve processes and customer service and to create new information services.
Verizon, then, will be relieved to have gained a major role in GE’s Industrial Internet ecosystem, in which AT&T was previously the only carrier partner. The telco is now working with GE to enable all the latter’s internet-enabled machines and devices to connect to Verizon’s M2M and cloud platforms.
However, Verizon has not, as yet, joined the IIC, which is going beyond GE’s own requirements and seeking to turn its Industrial Internet platform, developed for internal use, into a de facto standard for all kinds of players to use. The key partners in that effort are AT&T, Cisco, IBM and Intel, though it now has 80 members including Samsung and Microsoft. While the AT&T network will still be the default for GE’s network, it needs its connected industrial devices to work with any major M2M platform to support the full range of partners and customers and will look to extend its reach by partnering with a variety of telcos.
So Verizon will initially provide connectivity and services for users of GE’s own Predix
platform, which is the basis of the IIC initiative and which GE hopes will become a de facto standard for IoT connectivity when it opens it up next year. That will support applications such as remote monitoring, diagnostics and maintenance on a global level over Verizon’s wireline and wireless networks. The two companies are also exploring the possibility of creating a single global SIM for global connectivity, something already offered for M2M by AT&T.
“The potential for transforming industries – including rail, aviation, energy and healthcare, as well as society as we know it is tremendous, and yet the internet of things is a nascent, complex and fragmented market,” said Mark Bartolomeo, head of IoT Connected Solutions at Verizon. “Driving adoption requires broadening alliances across the ecosystem. We look forward to using the power of our network and cloud platforms to enhance our long-standing relationship with GE so that together we can create new business models across the Industrial Internet to meet the increasingly complex demands of customers.”
Verizon’s support for Predix highlights the way that GE will push its inhouse platform into every area of the industrial value chain, initially in the US and then internationally. It said last week that IIC members Intel and Cisco were now working to enable any carrier to integrate Predix into their existing network infrastructure, broadening the range of wide area network options for M2M users, and making it easier for operators to enter this market without developing significant new industrial connectivity platforms themselves. Intel is cooperating with GE to make devices and sensor networks ‘Predix-ready’ and it is integrating the capability into its IoT gateways along with features like enhanced security. The Predix software will also be incorporated in Cisco’s next generation of ruggedized network hardware for industrial environments.
This is not just about efficiency, or the chance to shape the industrial side of the IoT, as GE seems to be doing so far. It is also about hard cash. GE is well aware that, in time, there may be more revenue in the data all its pieces of metal generate, once they are connected, than in the machines themselves (and a machine such as an aircraft engine generates far more data than a smart thermostat). That explained last year’s $105m investment in EMC’s Pivotal big data division, and the creation of the IIC.
The US giant said it was on track to generate over $1bn in incremental revenue this year from more than 40 Industrial Internet offerings to customers and partners, and has clocked up $1.3bn in orders.
While Intel, Cisco and others race to incorporate Predix in their products, there will be even more significant IIC power to be had above that level, in the cloud where all the IIC data will be processed and monetized. Here, IBM will be the key partner, and GE’s own technology contribution will revolve around Pivotal. In the summer, GE announced that, with Pivotal, it had created a “data lake”, highlighting the magnitude of scale which emerges once the IoT moves from consumers to big industry.
Apparently, all the sensors in a GE jet engine, and elsewhere in the plane, will generate about 14Gbytes of data on an average commercial US flight. In its trials of its lake, GE gathered information from 15,000 flights on 25 airlines and fed all that, as raw data, into Pivotal’s PaaS (platform as a service), which is based on the open source data analytics platform Hadoop, and into Predix. In the aircraft trials, GE said it reduced the time to analyze all that data from months to days using the Pivotal/Predix combination, and that enabled one airline partner to reduce its fuel costs by 1%, thanks to better data on how plane systems were running (major US airlines spent $46bn on jet fuel in 2013, so 1% is not as trivial as it first sounds).
Bill Ruh, VP of GE Software and leader of the data lake projects, said the overall goal is to use machine data analytics on a vast scale to support better predictive maintenance for GE and its customers, and “get away from that alarm fatigue mentality – we want to know when a part is likely to break and watch usage patterns to see how parts can be more efficient and optimized.”
And of course, the same principles can be applied to other sectors. By next year, GE says it hopes to be collecting data on 10m flights a year, and to be sharing its data lake experiences and approaches with other industries, as well as applying them to its own customers and partners in many sectors. Its healthcare division, which makes machines such as CAT scanners, is now rolling out the technology, and next to follow will be the power generator, oil and gas, and rail and transport groups. On the technical agenda, as well as the IIC, will be increased integration of Predix and Pivotal.
While IBM competes with EMC and Pivotal in some areas, in the IIC it will be looking to work with the GE-backed platform to ensure its own cloud services take a prominent role. However, it is also pushing its own Bluemix PaaS elsewhere in the IoT, and has launched its ‘IoT Foundation’ to act as a repository for all the device data generated by the system, and to provide a secure way to share that data between services. Bluemix aims to rival Amazon Web Services as a developer-friendly cloud, and it runs on IBM’s SoftLayer cloud computing platform. Now the IoT foundation will enable connected devices to send encrypted data to the Bluemix database where it will be linked to other services via APIs, and stored. Developers will then be able to access that data and, if permitted, control devices which they do not manufacture.
IBM will form a partner program for the Foundation, and issue certified guidelines, or ‘recipes’, for connecting devices, sensors and gateways from those partners, which are likely to include existing IoT cloud allies such as ARM and Intel plus Elecsys, Multi-Tech Systems and Texas Instruments.
It is conceivable that IBM could adapt its repository to work with Pivotal and the IIC, and this market will enable the giant to harness many of its R&D efforts to create a broad-based platform and ‘as a service’ offering for the IoT. Obvious examples include the Watson cognitive computing engine, and the Node-red graphical user interface for programming interactions between things. IBM also developed the leading messaging layer standard for the IoT, MQTT (also part of the IIC platform), and is reportedly exploring using the blockchain technology behind Bitcoin as a device-to-device mesh. That could see the company extending its tentacles well beyond its database and cloud heartland towards the physical edge of the IoT – the mesh of connections which will link the objects to one another and to a longer range network like cellular or WiFi; and the all-important home or enterprise gateway, which will enable a supplier to gain control, and an element of lock-in, of a customer site. IBM is said to be working on highly intelligent gateways which will link back to its big data and cloud platforms.
It certainly won’t have this game all to itself. Many firms are building platforms to store, secure and interlink IoT data and services, including Salesforce.com and Philips in the medical arena (see separate item). But IBM does have the advantage of its broad reach of inventions and partners. “This packages together a lot of the technology we have been using, and it packages them in a new way and made them simpler because they are behind a managed service,” wrote Michael Curry, VP of IoT product development at IBM, in a blog. “Under the covers we are using the technology we have delivered over the years, but in a new, cloud-native deployment.”
“At the end of the day, the IoT is largely about data — the amount of data that the new world of connected things has created,” Curry said. “Today there are a lot of autonomous IoT devices but most companies don’t know how to harness that data. We are trying to help them do that, to simplify the process and to get value from that data.”
That is a goal which IBM will share with a host of enterprise, operator and vertical market giants, all desperate to turn the deluge of IoT devices and data into gold.