Internet of Things gateway: Q&A with Ido Sarig, Vice President, IoT Solutions Group, Wind River

March 1, 2013 OpenSystems Media

1What does the Internet of Things (IoT) mean for enterprise cost-optimization? Ido Sarig, Vice President, Internet of Things Solutions Group, Wind River explains how predictive maintenance and asset monitoring with IoT gateways enables an extra kicker for goods and services.



How does the Intelligent Device Platform relate to Internet of Things (IoT) gateways?

IDP will act as the gateway that sends data into the cloud. I think the gateway is a very strategic chokepoint. Although it might not be the area that generates the most revenue for Wind River and Intel, it is one that’s very critical because if you control the gateway to a large extent you define which sockets are going to win on the device side, and it will be those sockets that can send data to your gateway. You also control which cloud backend system is going to be the dominant one because it will be the one that is served by the most devices connected through your gateways.

IDP itself doesn’t do any of the picking of which data is to be picked up and sent forward. This would be application logic that would be up to the end user to define in the code. What IDP provides is a very broad range of protocols that enable you to collect the data from the devices using whatever means is most appropriate for your particular use case – whether wired or wireless, 2G, 3G, or LTE, wired Ethernet, Personal Area Networks (PANs), LoWPANs, ZigBee – whatever is suitable for your use case and whatever might already be implemented for your devices we will support, and that I think addresses one of the major challenges that companies looking to play in the IoT face, which is that there is no one single standard, no one-size-fits-all protocol that handles connectivity.

As far as selecting which data, one of the components of IDP is a very comprehensive application development environment that enables developers working on top of IDP to build their applications in the languages and environments that they’re familiar with. So we support Java development with OpenJDK, we support scripting with Lua, and we have basic C-level binding, so we provide a very comprehensive range of Integrated Development Environments (IDEs) for developers to work in. And they’re also very easily customizable because they’re all running today on top of our Wind River Linux, so if you’ve got other environments that you’d rather work with, you can easily extend it.

Do you think there is going to be a move towards standardization?

I think there is certainly going to be some level of standardization because that’s the nature of markets – you don’t want to be working with hundreds of different protocols, but I wouldn’t go as far as to say that everything is going to be a single protocol or everything is going to be IP-based moving forward – that I think would be wishful thinking. But I do think there is going to be some move towards standardization; I think we’re seeing customers react very negatively to the notion of a new solution that uses yet another proprietary protocol. There is going to be a drive towards standards – people want to use standards-based systems wherever possible.

Companies that want to stick with their proprietary protocols are certainly taking on some risk. They will be stranded and unable to interoperate with generic systems or other clouds. It’s a risk that they’re taking. Sometimes there’s justification for that; I’ve seen some protocols that have been highly optimized for round trip time or for determinism in round trip time, and there might be applications where the need to have this level of determinism or this level of response time overrides the risk of becoming an island. For most applications I don’t think that’s going to be the case.

How does Wind River work with its industry partners to facilitate application interoperability across various industry domains?

I can illustrate that with an example. One of our recent wins with IDP was in the retail sector in fiscal compliance. In a number of countries, the issue of ensuring that all transactions conducted at the point-of-sale are fully accounted for from a tax perspective is a very acute one. There is concern in the governments of those countries that there is a large-scale black economy that goes underreported, and they’re of course trying to limit the extent of that. So we worked with one of these national government’s tax authority who wanted to connect all intelligent point-of-sale systems in the country to a centralized compute system that is run by the tax authority to ensure that every retail sale is taxed appropriately. One of the things we were able to come up with was, on one hand, the IDP solution that provides the secure connectivity, but it also works together with the manufacturers or suppliers of these point-of-sale systems to ensure that their systems will be able to build on our IDP solution to connect to the Internet. It also worked with yet a third partner that provides the specific fiscal compliance software application built according to specifications of the tax authority that runs on top of IDP. So we certainly like to play a role of the lingua franca between all these entities, and our solution supports both the needs of the tax authority, the Original Equipment Manufacturers (OEMs) that manufacture the devices that need to comply with the regulations, as well as with software partners that provide the market-specific or vertical-specific applications.

We obviously provide ongoing support for all of these. In many cases there will be software updates that are required from our end to support additional new requirements or new protocols that might come along. But with the cloud backend that we’re developing we actually hope to be playing a bigger part down the road beyond just the connectivity solutions provided by IDP, and to work alongside these customers for many years to come – providing them with the basic capabilities of a Platform-as-a-Service (PaaS) to enable more efficient data flow from the devices into the backend, and then on the backend to provide facilities to collect the data, store the data, and analyze the data.

Can you give any details about your plans in the cloud?

At a high level I can tell you that this is a solution that is planned to be rolled out later this year. It is going be fully integrated with IDP in the sense that any IDP-enabled device will be able to seamlessly connect to this cloud backend. Some of the capabilities of this cloud backend are going to be storage of large data sets in a Hadoop cluster, the ability to act on the data that is received from the devices with a rules engine that will tie device data with business logic, and possibly close the loop with actuators back on the device side to take automatic corrective action.

In addition, we hope to provide more comprehensive data crunching based on machine-learning algorithms that will crunch data that comes from large numbers of similar or identical types of devices to address use cases such as predictive maintenance.

The addition of cloud solutions will enable updates throughout the entire IoT chain, correct?

That’s going to be one of the first use cases that will be deployed – that of software or firmware updates. I think that in many respects represents low-hanging fruit. A lot of the devices out there today require software updates or firmware updates on a fairly regular schedule. Unfortunately, today when that has to happen it often requires you to bring the asset into some workshop to have that work done, and if it’s not a mobile asset it requires sending field force engineers with a USB stick to physically visit the asset and update it. I think the opportunity to cut costs using remote software or remote firmware updates is enormous.

What do you suggest for companies that have not yet completely invested in IoT?

There are two areas of interest for companies to do this. Most start as what we would describe as cost-reductive initiatives; the lowering of maintenance costs of existing assets that are in the field would fall into this category. But I think once you’ve achieved the connectivity for the purpose of reducing your ongoing maintenance costs, you’re now in a position to move into the next category, which is innovative new revenue streams. One of our customers is a diversified manufacturer of electronic components that go into a wide variety of applications, one of which is home medical devices. This customer is seeing their margins on the hardware business slowly being eroded by low-cost competition from the Far East, and they’re looking to move upstream and monetize not just the hardware boards where the margins are low, but the value of data that’s associated with their boards.

One of their devices is a home medical device that helps patients recover from stroke. The way this device works today is that it stores all the session information on an SD card and periodically the patient has to come in for an appointment with his physician. The physician extracts the SD card from the device, puts it into his laptop, reads all the parameters, and then decides how to further tweak the treatment parameters. So what they want to do is connect this device to an IDP gateway that would reside in the patient’s home – or maybe if we’re talking about a large clinic it would be multiple devices connected to a single gateway – and send the data online to the physician, enabling him to read the session data online without having to schedule a physical appointment, greatly improving the productivity of the physician who can now serve multiple patients in one sitting instead of having to schedule an individual appointment with each one of them.

So far this is a cost reductive value proposition, but here’s the kicker. Now that they’ve got all this data collected for multiple devices, you add on the ability to anonymize the details so they’re not associated with an individual patient, and you can now go back to the insurance companies that are paying for this treatment and offer them for-a-fee access to the data that will enable them to analyze the effectiveness of the treatment, and then stop payment as soon as it reaches a plateau of diminishing returns.

Wind River Systems

www.windriver.com/solutions/m2m/

@WindRiver

YouTube: www.youtube.com/user/windriverchannel

Blogs: blogs.windriver.com/

Brandon Lewis (Assistant Managing Editor)
Previous Article
7 rules for designing wearable devices
7 rules for designing wearable devices

The past year has certainly been an exciting time for the semiconductor industry. The shared enthusiasm amo...

Next Article
Effective hardware-software co-design for automotive systems
Effective hardware-software co-design for automotive systems

Cadence Design System expert Frank Schirmeister explains how hardware/software co-design can aid SoC develo...