Late to the party or leader of the pack? IBM blends lessons from Smarter Planet, Rational, and Continuous Engineering in new IoT division


In March IBM announced its intentions to invest more than $3 billion over the next four years into a new Internet of Things (IoT) business unit they hope will empower developers and foster ecosystem growth in the world of interconnected everything. But are they too late?

In this interview with Amit Fisher, CTO of IoT Continuous engineering at IBM, he explains how lessons learned from the company’s Smarter Planet, Smarter Cities, Rational, and Continuous Engineering initiatives of the past have given them a head start as they rollout their next generation of innovation enablers for the IoT.

Earlier this year IBM announced a new IoT business unit. Can you describe this new division and your role within it?

FISHER: Multiple business units are coming together to form this new unit, so it’s very big. The reason this was constructed out of multiple divisions is that the overall goal is to support IBM’s industry solutions to create more connected systems. So this is a direct continuation of our Smarter Planet play. If you take a look at Smarter Planet and Smarter Cities, this is exactly the topic that we have been talking about for a long time. Now the new term that’s being used in the industry to describe it is the Internet of Things (IoT), but it’s really the same thing.

I’m the CTO of IoT Continuous Engineering, and if I could describe my role in a single sentence it would be, “How we help customers create smarter IoT systems and how we make them successful as they build their IoT systems.” That relates to the overall lifecycle of building an IoT solution, from understanding your requirements and understanding what you’re trying to achieve to the design of the system, which includes all the components of the system such as the device itself, cloud-based services, the client side, and especially the overall architecture that considers both the structure and the behavior of such a complex system. This is a huge challenge as opposed to designing an aircraft or a car in the traditional way, not that it was easy, but because the problem was self-contained. It was well defined what you were trying to achieve, and while it was very complex and hard to do, you could control most of the variables in the system. In an IoT system you are working under the open world assumption, so by definition you do not control anything and you need to design a very adaptable system that will work in ways you can’t necessarily predict on day one of the project. It’s a much more complex environment and much more of a design challenge.

The idea behind these systems is that connectivity will change everything that we know about complex systems. Connectivity will have a central position in the design of everything that we do, which means that the infrastructure is going to be different. The types of services that have been released and the types of analysis you can do today based on data that is available were not around five or ten years ago. The concepts are not new because people always wanted to use data for analytics, but right now data is relatively cheap to get – it’s not necessarily easy to analyze and get insight out of it, but it’s relatively cheap to get – so the opportunities are much bigger and being created at a much faster pace.

It may seem that IBM is a little late to the party with the announcement of an IoT business unit this year, but, as you mentioned, IBM’s Smarter Cities and Smarter Planet initiatives have been underway for some time. How will the elements of those programs integrate with this new business unit?

FISHER: The thing that we’ve learned over the last few years in developing Smarter Planet solutions is the real insight you need to bring products to market. The fact that today we lead the analytics space and the fact that IBM leads the Industrial Internet space is because of that Smarter Planet strategy. Again, many of the things that are now described under the context of the IoT have been there, and IBM has been using the terms “interconnected” and “intelligent” to describe the patterns that we saw in Smarter Cities; if you take a look at all the buzz around IoT and Big Data, it’s the same thing. It’s all about the fact that now data is available and organizations will know how to make use of that data in a smart way that allows them to leapfrog the competition.

We’ve learned a lot in general, and the lessons learned are about how you collect data and what are the right ways to maintain data. For example, predictability and predictive maintenance may be the most known example of how to maintain equipment not necessarily by the spec or when things stop working, but by using more sophistication in general, more analytics that allow me to predict when a system might fail to save cost before something much worse happens. So there are a lot of implications on the operations side, but there are also a lot of implications on the design side. We’ve been working with clients under the Smarter Planet umbrella, and when we took a look at operational metrics we were able to propagate that back to the R&D department so that engineers could make better decisions when designing their next versions. So yes, we’re definitely using a lot of the lessons and patterns we have seen in Smarter Planets and Smarter Cities solutions to develop our IoT agenda.

As far as product and services offerings, what can we expect from the new IoT business unit, both on the whole and within the Continuous Engineering division?

FISHER: In the broader IoT division we are introducing a lot of new capabilities. We are introducing an infrastructure for IoT connectivity, which is called IoT Foundation, and we are releasing data analytics services that allow the very, very easy creation of smart analytics and smart widgets on top of our Bluemix platform to better understand and analyze your data. So that’s the broader sense of the IoT division.

Within the IoT Continuous Engineering environment we are also looking at new areas and new capabilities that will allow customers to innovate much faster. Earlier this month we actually released an experimental service capability we call IoT Workbench that we are very excited about (Figure 1). The IoT Workbench is basically a rapid application development environment that allows you to very quickly define, design, and later on simulate an end-to-end IoT solution. So it allows you to easily define the classifications of your design, the devices you are going to use, and the cloud-based services that you want to use in order to create a business project or the application. We have integration, for example, with Node-RED that allows you to easily define the context or logic of cloud-based services, as well as the client side that can file events or file commands into the system. Very, very easily – and by easy I mean in a matter of minutes – you can define an architecture and you can start playing with the implications of different design decisions that you make, including the ability to simulate the whole environment on top of our IoT Foundation cloud with the touch of a button.

Figure 1: IBM’s IoT Workbench is a platform that can be used to design, build, and simulate IoT devices, cloud services, and clients before the devices or solutions themselves exist. Currently offered as an experimental service, the IoT Workbench can be accessed for free at
(Click graphic to zoom by 1.9x)

We are very adamant that this is an experimental service and not ready for primetime, but we have released it as a minimum viable product to let the market know where we are going and what we are looking at. We are very excited because we think that part of the challenge, in any solution and specifically in IoT solutions, is that today’s engineers need to be experts in many different fields. If you think about the typical IoT solution, you need to understand the device, you need to understand the embedded logic of the device, you need to understand the different protocols that exist in IoT, you need to be a cloud expert because most of the logic today and a significant part of the analytics is done in the cloud, and you also need to be a client or mobile expert because most IoT solutions translate eventually to some type of client or user interaction through either a mobile device or web browser. So there are many different technologies and pieces that you need to learn in order to develop those types of solutions. What we’re trying to do with the IoT Workbench is quickly prototype and analyze how the system will behave in a virtual environment that allows designers to make more informed decisions very early in the product lifecycle, before the final solution exists or before the device even exists. This integrates with the other technologies and products that we have developed, so the idea is that there will be seamless integration with IBM Rhapsody, our mature model-based engineering environment that allows you to further refine the logic on target devices. It will also integrate with our requirements management tool and change management processes. So it’s not a standalone platform but rather extends capabilities to the IoT developer.

Based on the tools and capabilities you just described, do you see the future role of vendors like IBM centering around platform-as-a-service (PaaS) offerings for developers looking to build IoT solutions?

FISHER: Yes, absolutely. That’s exactly where we’re going with Bluemix. We want to make Bluemix a landing page for every developer, every designer, every engineer, every system integrator that is developing something. Our responsibility is to make it very efficient, to make them productive from day zero, and to offer the types of services they need to construct their solution without having to reinvent the wheel every time they need to build something. Specifically in Continuous Engineering, this is how we have worked for many years – how you provide a platform that allows engineers to focus on the innovation and not on the planning or the accidental complexity of integration. We want to make their life easier so developers can create much smarter solutions.

There’s recently been a lot of talk about the IoT being a bubble about to burst. What are your thoughts on the IoT bubble and its prospects over the next few years?

FISHER: If you look at the term “IoT” it’s definitely now at the peak of its hype, so there will be some realizations. But IoT is not a bubble because some of those realizations have been in technology that is here, is happening, and is here to stay. So, regardless if you believe that billions or trillions of IoT devices will be out there, I don’t really think those numbers are important. What is important is that we’re already seeing value, business models, and revenue models that are changing because of those types of technology. Again, those things are not new concepts – wireless networks and sensors have been here for decades – but what has changed is the fact that now it’s very, very cheap and very, very easy to do it under a set of standard protocols that allow you to do it very fast. Two guys and a dog, as they say, can create an IoT solution in a matter of weeks, maybe a few months, and a pretty mature one. That was not possible a few years ago. So it’s here to stay.

I think some of the realization will be that a lot of hype on the consumer side will translate into tangible solutions that bring value on the industrial side, or what is called the Industrial Internet. We will see more and more traditional manufacturers take advantage of IoT, not because of the hype, but just because it’s good business.

Amit Fisher is CTO of the Internet of Things Continuous Engineering group at IBM.

IBM Internet of Things Continuous Engineering