Gallery: 10 Questions to Ask When Selecting an IIoT Platform
It can be hard to wrap your head around the IIoT platform market, but after answering these questions, you should have a clearer sense of priorities when shopping for such software.
September 13, 2019
![Image shows the inside of a smart factory Image shows the inside of a smart factory](https://eu-images.contentstack.com/v3/assets/blt31d6b0704ba96e9d/blt9a761f533ad3c845/63abedfc25035157f3d99603/GettyImages-1023232352.jpg?width=700&auto=webp&quality=80&disable=upscale)
Getty images
This question may sound simple, but it is one of the toughest to answer well. Industrial companies should not limit their long-term digital strategy to rough-hewn objectives such as improving operational efficiency or cutting costs — vital though those goals are. While IoT technology can enable a bridging of the gap between the physical and digital worlds that can inspire new business models in the long-run, a significant number of manufacturing companies encounter delays in their present-day IoT deployments. Our 2017 survey revealed that 55 percent of manufacturers reported their IoT projects took more time than expected.
Given the real risk of encountering unexpected delays when deploying or scaling an IoT project, industrial companies should ensure that the IoT platform technology they select meets their short- and long-term needs and that the vendor and its partner ecosystem can further support those demands.
It is vital to partner with IIoT platform providers that have a solid customer base and a robust ecosystem of partners. A successful platform vendor should be willing to provide a clear sense of the installed base of the platform, references and a clear sense of how their customers are using their platform. Furthermore, it is important to determine the reliability of the partners that form the vendor’s ecosystem. While a federated ecosystem of partners is more likely to meet the needs of specialized manufacturers, it can be difficult to orchestrate an alliance of IIoT platform partners. It is therefore important to have a clear strategy and determination of the IIoT project ownership.
There is no such thing as a one-size-fits-all industrial-grade IoT platform. “You want to make sure that whatever platform you decide to move to is aligned with your industry,” Kocher said. “The platform may not be tuned specifically for the vertical you’re in, but it certainly needs to be aligned with your industry.” That means the broader ecosystem surrounding the platform should have expertise in your industry and other players from your industry niche should have experience with the platform in question. “That’s going to have a huge impact because those may be partners with you or your customers may be using them,” Kocher said. Related themes to investigate include the vendor ecosystem’s familiarity with core industrial use cases such as track and trace, predictive maintenance, field service management, environmental monitoring and so forth.
As internet pioneer and venture capitalist Marc Andreessen explained in a 2007 essay, a platform is “a system that can be programmed and therefore customized by outside developers.” In other words, users can customize platforms. A platform can thus be “adapted to countless needs and niches that the platform’s original developers could not have possibly contemplated, much less had time to accommodate.” By Andreessen’s definition, if a self-billed platform can’t be programmed with code qualified by the user, it doesn’t qualify as a platform.
While a platform’s ability to help support core “plumbing-type” connectivity is important, the greatest value of an IIoT platform lies in its potential to provide innovation from a third-party ecosystem that can supplement the efforts of an organization’s internal R&D efforts. When selecting a developer-centric platform, it is therefore important to determine how many users are actively contributing to a vendor’s platform ecosystem and to further understand what the vendor is doing to recruit and retain developers.
This simple-sounding question can be difficult to answer given the crowded IoT platform marketplace and the variety of pricing and licensing models. Prominent tech vendors provide their pricing model online. While some models are straightforward, one prominent cloud company has a more-complicated pricing model with variable device provisioning costs based on separate location costs for fleet indexing and search, as well as device jobs such as orchestrating a firmware update to remote devices. Depending on your organization’s needs, platform costs can vary significantly. In one hypothetical scenario, the monthly costs of an IoT platform from hyperscale cloud vendors ranged from as low as $250 to as high as $1,382.40. Given the complexity involved, industrial companies should come up with a limited list of, say, four or five scenarios and perform an in-depth calculation of what each of them would cost and estimate their potential to support your most important business outcomes.
Once your firm has whittled down IoT platform companies to a manageable list, the next step is comparing the final contenders against one another in a head-to-head test before taking the results to executives for final approval. In the first phase, your firm should determine what the integration path looks like between your current systems and the platform you ultimately select. “You can have a great new platform, but it might force you to drop everything and rebuild much of the existing software you use from scratch. That would be painful,” Kocher said. “If there is an integration path or migration path, that’s a huge benefit to you. Some of that falls into the supportability and training. Nothing’s going to be perfect. But if there’s, you know, some customization and consulting to help you get through that migration, so much the better.”
Cybersecurity remains one of the most common barriers to IoT projects in general. In manufacturing environments, IoT endpoints can capture sensitive information that is potentially attractive to black hat criminals interested in either espionage or sabotage. Last year, EEF’s Cyber Security for Manufacturing report concluded that half of the respondents had to deal with a cyberattack. Among those who were hit, half of those reported their firm had an adverse business impact. Worse, such attacks are on the rise as manufacturers across the globe contend with an upwelling of cyberattacks. The National Center for Manufacturing Sciences in the United States estimates the cost of such breaches range between $1 million and $10 million. To deal with such concerns, IoT platform providers are gradually adding more end-to-end security features to their products.
While the initial focus of IIoT platforms was to provide middleware that ties together IoT devices and gateways, data analytics and machine learning are important considerations when selecting an IIoT platform. Gartner anticipates that by 2022 on-premise IIoT platforms in conjunction with edge computing will drive 60 percent of IIoT-based analytics.
A successful IIoT platform with analytics capabilities can help industrial companies transform how they traditionally process data, helping improve how they visualize, contextualize and analyze data streaming from a variety of sources. Traditionally, the legacy analytics deployed in industrial settings were limited to relatively low volumes of data. Oftentimes, that data was siloed and decentralized.
It is therefore important to determine how well an IIoT platform can ingest and synthesize large amounts of data of varying types.
When assessing a vendor’s degree of support, your organization should evaluate not just the level of customer support it offers in your region, but the availability of end-user training, documentation and the level of peer resources available. How frequently does the firm tend to offer updates to its platform? Does the vendor support over-the-air software updates for your IoT project? Again, another important consideration here is the platform’s potential life span. Even if the vendor provides excellent support now, you want to be confident that they continue to do so in the future. It also may be the case that the vendor has a great platform with impressive technology and APIs, but doesn’t support it very well. “If that happens, suddenly that burden falls on you where you’re going to spend a lot longer developing things,” Kocher said. Conversely, an IoT platform vendor with a robust developer community gives you a plethora of options for software development to meet your firm’s specific needs. “You are likely to be able to find someone who can build custom software for you, or who can customize existing code,” Kocher added. “Or their documentation and training can help your internal development team accomplish the same objective.” Finally, it should be clear who at the vendor will take responsibility of a problem, should one occur, according to Steve Brumer, business partner at IoT Group. It should also be clear what the vendor’s — or its partners’ — availability is like for contract development work if needed down the road.
The roots of IIoT in factories and other industrial environments stretch back decades, encompassing machine-to-machine and networked automation technologies such as Supervisory Control and Data Acquisition. But until recently, it has remained difficult to obtain a holistic view that integrates data across the various sensors, data types and computing hardware deployed in an industrial environment. The Internet of Things promises to give manufacturers a single digital thread that weaves together data flows from across your industrial environment. So the last thing you want is to deploy an IoT platform that ends up being something of a walled garden that attempts to lock you into using only products from the platform provider. The IoT platform your firm selects should ideally support other software platforms — potentially in tangential industries. “It may be that you’re really focused on one industry, but 20 percent of your customers are in a related industry and are using another platform. How does the platform you’re going to tie into communicate with that?” Kocher asked. “Are there API developer kits and programs to allow people to do that?”
The roots of IIoT in factories and other industrial environments stretch back decades, encompassing machine-to-machine and networked automation technologies such as Supervisory Control and Data Acquisition. But until recently, it has remained difficult to obtain a holistic view that integrates data across the various sensors, data types and computing hardware deployed in an industrial environment. The Internet of Things promises to give manufacturers a single digital thread that weaves together data flows from across your industrial environment. So the last thing you want is to deploy an IoT platform that ends up being something of a walled garden that attempts to lock you into using only products from the platform provider. The IoT platform your firm selects should ideally support other software platforms — potentially in tangential industries. “It may be that you’re really focused on one industry, but 20 percent of your customers are in a related industry and are using another platform. How does the platform you’re going to tie into communicate with that?” Kocher asked. “Are there API developer kits and programs to allow people to do that?”
In prominent platform markets in the past, cloud computing and connectivity serve as the foundation for co-creation. The platform helps solve a central problem while developers build on the platform to meet custom market demands. Smartphone ecosystems, for instance, are an example of this, where Google’s Android ecosystem and Apple’s App Store both inspired a symbiosis of internationally-based developers, consumers and entrepreneurs. In 2018, the App Store generated $22.6 billion in international app revenue while the Android-focused Google Play store generated $11.8 billion. Companies ranging from Amazon to Google to Facebook to Salesforce have all created platforms that support larger ecosystems.
While the IoT platform market has yet to replicate the success of those prominent examples, executives who are mulling using an IoT platform or multiple platforms should strive to select a platform that not just simplifies commodity-based “plumbing” tasks, but can support an ecosystem of innovation that helps your firm make tangible improvements in how it manages industrial assets and operations. Here, we propose a series of questions designed to do just that.
About the Author(s)
You May Also Like