What Engineers Need to Know about Communication Protocols When Choosing IoT Management Software

A Failure to Communicate on the IoT

Engineers can design the best product imaginable and let it connect to the Internet of things (IoT). However, if it doesn’t play well with other products, it will be doomed in the marketplace.

According to a report by PlumChoice, two-thirds of all those who try to use IoT devices will be thwarted for one reason or another. One out of every seven who have bought an IoT device have returned it—and over one-half of those said they would never buy that brand again.

“Despite the significant opportunities presented by innovative connected home and IoT devices, companies are failing to help consumers integrate these products into their already-complex technical environments,” said Steve Thompson, vice president of IoT business development at PlumChoice.

The promise of IoT comes to life as things connect to each other. However, Paul Didier, AVnu Alliance board of director and IoT solution architect at Cisco, explains that many IoT devices remain segmented on highly non-connected networks due to incompatible protocols as well as proprietary reasons. “With IoT, a lot of the core system device applications are impossible with this segmentation, such as big-data analytics,” he said. “Convergence is a key idea.”

To connect these segmented IoT networks, engineers will typically need some form of overarching management software to bring all of the things together. However, what should be simple IoT integration isn’t as simple as it seems.

“Without the necessary pre-purchase and ongoing support, ‘thing’ makers are looking at a long road ahead that involves a lot of unsatisfied customers and many no-fault-found returns,” said Thompson. “We need to help people derive real value from these investments—and, in return, brands will be more successful.”

A Networks of Networks

“IoT management software encompasses a lot of complicated capabilities,” said Bryan Kester, director of IoT at Autodesk. “Most of these are ideal to have, but you don’t have to have them all.”

When it comes to actually implementing IoT, the engineer is faced with a dog’s breakfast of various concepts, tools and technologies. These include—but is not limited to:

  1. Remote firmware and firmware management software for collecting data
  2. IoT device monitoring software for assessing operation of IoT devices
  3. Data storage over the cloud or servers
  4. Big-data analytics for assessing and optimizing a system
  5. Version control for pushing updates to the IoT device

“IoT control is very broad,” added Todd Walter, AVnu Alliance chair and senior group manager at National Instruments. “There are some common denominators between management software, Web services and protocols. They have management concepts for end nodes and sizable communities that support them. But there is no single management protocol today—nothing that does it all. I think it will depend on the market to see if one is needed.”

Microsoft Azure IoT architecture. (Image courtesy of Microsoft.)

As previously mentioned, many IoT products will have their own proprietary software that connects the device to the Internet, manages its data and controls the product. Didier explains that, when compatible, IoT management software adds another layer of communication on top of these detached networks.

This overarching network of networks will still let each proprietary software control their things. The main difference is that this network of networks will allow for communication protocols for big data, optimization and system control opportunities.

Amazon, Microsoft and PTC Step Up

For example, Microsoft Azure and Amazon Web Services (AWS) look to perform this connectivity and management through their cloud platforms. Communication with the hub is done over HTTPS and MQTT protocols and custom gateways when needed. Microsoft Azure can also use AMQP protocols. However, if your device isn’t compatible with these protocols, then they will not be compatible with Azure or AWS.

How AWS sees IoT. (Image courtesy of AWS.)
“You deploy your event hub, add your devices and connect your varying devices through the hub,” said Ken Patterson, technical solution specialist for Microsoft Azure at N3. “There are third-party tools you can utilize like Raspberry Pi. You can then use development environments and Arduino to connect to the Raspberry Pi. As a result, your mobile platforms on Apple, Android and Windows can all integrate and be used for control or as a sensor.”

Another example of a leading IoT management software is PTC’s ThingWorx. ThingWorx is positioning itself as a protocol-agnostic choice for IoT management software. “At the core of the ThingWorx platform is a comprehensive modeling environment that also generates REST APIs for the ‘things’ in the model,” said Joe Biron, vice president of IoT technology at ThingWorx. “The REST API not only allows for the management of the IoT devices, but also allows for communication between the disparate cloud-based systems, enabling IoT applications to be extended to other domains such as CRM (Salesforce), cellular (Jasper) and so on.”

However, ThingWorx will not be as simple as you may expect. An engineer or IT professional will still be needed to implement the IoT system connections, create the dashboards and ensure the system is operating properly.

How to Choose an IoT Management Software?

Many of the top players attempting to integrate the IoT will have familiar names. This is because they are likely already a major player in another technical vertical.

The IoT management software options the engineer wishes their product will be compatible with will dictate the protocol selections for an IoT device. When making this decision, engineers should pay attention to the end users of their management software, which could be themselves, and the use case of the IoT connection. For a list of some IoT Management software options for engineers, see the table below.

Company

IoT Management Software IoT Product’s End User Popular IoT Use Cases
PTC ThingWorx Product’s Design Team, Industry Business Management, Big Data, Analytics, Product Optimization, Failure Management
Autodesk SeeControl Product’s Design Team, Industry Business Management, Big Data, Analytics, Product Optimization
Microsoft Azure IoT Hub Industry, Consumer Business Management, Big Data, Analytics, Failure Management, Consumer Products
AWS AWS IoT Industry, Consumer Business Management, Big Data, Analytics, Failure Management, Consumer Products
Google Weave Consumer Consumer Products
Apple HomeKit Consumer Consumer Products

For example, a product development team may want to keep track of how their products are interacting with their customers. This data can be used to improve the product design in the future. As a result, they might want to pick an IoT management software that has integration to design software such as PTC or Autodesk.

So how do you start to determine which IoT management software to use or who you can work with to create your own? Didier thinks a good place for many to start is with their previous IT partners.

PTC and Autodesk’s take on IoT

Engineers might want to see what IoT options their design software provider might offer. For instance, PTC has invested over $750 million dollars into IoT acquisitions. As a result, PTC has amassed a significant amount of IoT technology. Whereas Autodesk has recently launched their SeeControl product. As a result, a lot of news is sure to come out of both of these design companies and their competitors are sure to follow suit.

However, as no IoT management software is all encompassing, “there is sometimes a need to have two systems,” noted Kester. “Some of these offerings have open, cloud-available APIs, or such capabilities are on the way. However, there is no perfect IoT software vendor today, and some systems are built more for data sharing than others.”

How Does Potential IoT Protocol Standardization Affect My Management Software?

One of the leading reasons the choice of IoT management software is so important is that there is a lack of IoT standardization. As many product designers opt to create their own proprietary control software, the problem only gets worse.

“In the cloud, there is a lot of work being done to bring the data all together, but IoT isn’t as developed,” said Didier. “How to use different systems and applications from various different vendors hasn’t been done.”

However, there is hope for standardization in the IoT community, and it is taking on a few forms. Keeping track of the IoT standardization field can keep you ahead of the curve to ensure your IoT device will make the cut.

IoT Protocol Standardization Option: Industry Consortium

The first potential sources for IoT standardization are the industry consortiums. These groups ensure that all products made by members of the consortium are compatible. Some of the leading consortium and standards organizations include AVnu Alliance, Industrial Internet Consortium (IIC), ODVA, OPC and Z-Wave.

As a result, becoming a member of an industry consortium and ensuring your management software meets all of its protocols is a good way to maximize the number of things your product will be compatible with. Besides, their protocols will be a good starting point if you are opting to create your own proprietary management software.

Each consortium creates standard protocols that are optimized for different situations. If there are consortiums creating protocols that match your use cases, then you should look into how to become a member.

For example, AVnu Alliance is focused on ensuring time synchronization between nodes for the purposes of Industrial IoT control and analytics.

“What AVnu Alliance is doing is taking the IEEE standard for time-sensitive networks (TSNs) and making that operable and certifiable between devices,” clarified Didier. “This is similar to what the Wi-Fi Alliance did for 802.11 media access control. They created a network where you always get Wi-Fi connectivity despite the device you are using. We want to do the same for TSNs in industrial applications.”

Potential TSN architecture. Image courtesy of National Instruments.
Alhough no consortium has come up with a complete IoT standards solution, Didier notes that the IIC and Germany’s Industrie 4.0 have put in a considerable effort.

“Neither of them has settled on what the full architecture will look like or what the key technology will look like, but they have guidelines,” he explained. “Industrie 4.0 is looking toward TSNs to converge networks as well as open communication platforms to cross the data across large spaces and run the larger business processes. Additionally, IIC has published a paper outlining a reference architecture, including the key standards and technologies. But they both still need to outline the next steps, which will be done in 2016. AVnu Alliance will be helping both of them to ensure the convergence of industrial devices within these networks.”

IoT Protocol Standardization Option: Piggyback on Internet Standards

Because the things will all end up on the Internet eventually anyway, some organizations are looking to bring standardization using Internet protocols.

“In a lot of cases, the IoT devices are using standardized protocol to transmit data anyway,” Patterson explained. “As for standardizations for the IoT protocols themselves, a lot are being built to integrate, despite these differences. For example, the IoT protocol gateway can integrate multiple transmission protocols into multiple locations.”

Walter agrees that standardization can happen after the things connect to the Web. He suggests that Ethernet might be a solution. “One key move to convergence comes as these products move over standard Ethernet,” he said. “We can then tap into those standard Ethernet communications. We think this will help with the convergence of communication in Industrial IoT.”

How much IoT standardization is needed if they are going to be brought into a standardized location or communication protocol eventually? After all, Microsoft Azure and AWS already connect to multiple IoT devices using HTTPS and MQTT protocols.

IoT Protocol Standardization Option: Consumer and Market Pressures

Users want to have their IoT device work to their expectations out of the box. As a result, “you will see a shakeup of consumer friendly IoT interfaces,” said Chuck McDaniel at N3. “From a consumer viewpoint, they will want something easy in order to get those analytics.”

Unfortunately, the marketplace doesn’t always listen to what is best for the consumer. “Capitalism is working against any standards as well as the success of any governing body,” Kester said. “If you look at an open and ‘free’ standard like the Android operating system, you see it bifurcate and split in flavors across handset manufacturers, TV manufacturers and even industrial manufacturers that are adopting it for IoT controllers. They are free to modify it to their own unique purpose and have done so to keep competitor technologies off of their ‘things’ and to sell the apps and services they want on their device.”

This is an interesting point. When given a clear standard such as the Android operating system, and a demand from the consumer for compatibility, the market still decided to split the standard into many in order to suit their own proprietary needs.

“In the tech industry, companies are very used to being both competitors and allies, because they know they can make a bigger market together,” lamented Kester. “That is why there has been such dramatic progress and growth in tech over the last 20 years. Industrial manufacturers don't yet have this mentality.”

Since this method of IoT standardization seems to be more of a format war, engineers should pay close attention to how the format war’s pendulum might swinging.

IoT Protocol Standardization Option: A Standard IoT Chip

An alternate source for standardization suggested by Kester is at the chip level. “That is probably one of the best possible outcomes—IoT protocol choices are determined by the embedded system and its tools,” he said. “However, there are many different things with different microprocessors out there, and they all have constraints on how they operate.”

Not all IoT device hardware will have the same CPU power or memory. As a result, many of the low-powered options will need to be very minimalist in their IoT protocol or risk maxing out the chip. However, the chips with more computational capacity will need more power to operate. As a result, a one-size-fits-all solution for IoT standardization from the chip level has its own complication.

An interesting pair utilizing this strategy are Marvell and Google Weave, which recently released a series of plug-and-play IoT chipsets that are Google Weave-ready. Although this particular chip may not get a big enough push to conquer the market, it will certainly simplify the IoT product development cycle.

If there is a chipset that connects to a management software that meets your needs, it might reduce your time to market. But you should make sure this solution meets the needs of your end users and IoT use cases.

IoT Standardization—The Cost of Waiting

The longer it takes a design team to bring their products onto IoT, the cheaper and more polished the management software will become, especially as standardization develops into a growing reality.

This might look appealing, but it is a double-edged sword. First, standardization isn’t guaranteed. In addition, waiting to implement IoT has a heavy cost in opportunity.

“There is so much value you can get from the integration,” Walter said. “You may find though that by the time the dust settles, you are no longer competitive and it’s too late to get into the game.”