On June 2nd, IBM and Cisco announced a partnership for IoT analytics which aims to combine IBM's cognitive computing analytics in the cloud with Cisco's fog computing analytics at the edge. Overall, the combination is a positive one for IoT and analytics decisions makers since it begins to help clarify the IoT data journey story.
Why It Makes Sense for Cisco and IBM to combine efforts for IoT Analytics
As I mentioned in a previous blog post, it seems virtually everyone in the IoT ecosystem claims some sort of analytics capability. This makes perfect sense, since insights and information extracted from sensor data (and integrated with other data sources) is ultimately the source of value for IoT initiatives. However, it also creates confusion. There is a distinct lack of clarity in the market in terms of what type of analytics should be executed across different parts of an IoT system, as well as how different analytics technologies and tools can fit together. Quite frankly, IoT analytics seems — excuse the pun — kind of foggy. With its collaboration, IBM and Cisco are attempting to clear the air a bit.
The problem IBM and Cisco is attempting to solve is related to the challenge that many IoT deployments have: in certain environments (such as oil and gas, and other distributed, industrial systems), executing analytics at the edge is both necessary and practical. Practical since transmission costs for large amounts of data could be prohibitive, but also network availability/reliability to do so may be lacking. And necessary, since there may be time-sensitive analytics that need to be run on streaming data for mission-critical applications. At the same time, there is incredbile value at conducting advanced analytics to enhance understanding of operational performance or predictive maintance.
What IBM and Cisco Each Bring to the Table
Just to clarify, IBM and Cisco have not announced a joint product. Instead, they have created integration points between their two distinct product packages to ensure they work together, and combine different elements of their solutions. Each company will continue to sell its own products through its existing channel, and will not resell each other's products. However, the different capabilities are highly complementary. Let's take a look at what each company brings to the combined solution.
What Cisco brings:
- Edge and fog computing capabilities, which allows data to be processed locally.
- Edge analytics on streaming data for asset and performance monitoring based on defined rules.
- Ability to aggregate data at the edge or fog node, and send relevant data and/or summaries of data to the cloud for further processing.
What IBM brings:
- Advanced cloud-based analytics capabilities, such as cognitive, predictive, and machine learning.
- The ability to leverage cloud-based analytics of large, historical sources of data to develop rules — such as rules for specific application use-cases — and run analytics where it makes sense. For example, rules can be defined and then pushed out to an edge node or gateway/router to be executed locally.
- Integration with additional internal and external data sources in the cloud (such as IBM's own Weather Company data) to provide deeper analytics that can help further refine rules.
IoT Analytics Impact
The announcement seems like a no-brainer for both IBM and Cisco. Both are large, trusted enterprise brands, and joining forces only enhances their validity as IoT leaders. And as mentioned above, there are true complementary capabilities here, and the integration of these different product/solution approaches provides some clarity about how analytics can be executed across the entirety of an IoT architecture. Here are a few additional thoughts/observations:
- There is no one-size-fits-all approach to IoT system design, which includes analytics. IBM and Cisco used marketing terms such as seamless and out-of-the-box to describe their joint effort. However, IoT solutions are complex and highly custom. Creating integration points between the solutions make sense, as many technology components need to be combined — by integrators, resellers, enterprises themselves — as solutions are created and deployed.
- This is not an exclusive relationship. It wouldn't be a surprise to see more announcements from both of these companies as they seek to address different IoT solution requirements for different use cases and industry segments. There are many edge technologies and vendors into which IBM will want to plug its cloud-based analytics capabilities. Likewise, there are many cloud/data center-based analytics tools with which Cisco would be wise to develop compatibility. Again, this is not a one-size-fits-all proposition.
- It was interesting to see Bell Canada's Wade Oosterman as a guest during the webcast. After all, Bell Canada is a large network operator, and theoretically, edge analytics is designed to limit the amount of data sent over networks. However, this made sense on a couple of levels. First, it indicated that even operators such as Bell Canada recognize that not all IoT data can be sent over things such as mobile/cellular networks, and meeting customer requirements is critical for IoT success. Secondly, mobile operators represent a key partner to Cisco in it's IoT efforts (especially given it's acquisition of Jasper), and it was very compelling to have a carrier partner validate the approach.
- It was also interesting that IBM didn't mention Quarks — and that Quarks is not part of this solution. IBM Quarks is an open-source, lightweight runtime designed to be embedded in "things" to do edge analytics. Clearly, in terms of the Cisco-IBM relationship, the edge analytics capabilities are being left to Cisco. But it will be curious to see how IBM ultimately positions Quarks as part of a broader IoT analytics story.
As mentioned, overall this is a strong announcement for both IBM and Cisco, and a positive trend for the IoT market at large. Ecosystem partnerships will be crucial in driving IoT solution growth. These partnerships will require more than just vague mentions in marketing materials, and instead need to incorporate tighter product integration, which appears to be what Cisco and IBM are jointly pursuing. And finally, any clarification of the IoT data journey is a welcome one for buyers and decision makers.