IoT integration services aren’t quite what they use to be.
In the early days of IoT, projects had a science project vibe. Organizations faced the challenge of cobbling together multiple technologies — sensors, software, connectivity and other parts — from a diverse set of vendors. The service provider’s role was to attain basic IoT integration at a time when “turnkey” offerings didn’t exist.
Times have changed, however. IoT has matured since first appearing on the enterprise IT radar about a decade ago. Integration isn’t entirely DIY anymore. AWS, Google and Microsoft offer IoT services on their public cloud platforms, and IT companies offer out-of-the-box IoT functionality to jumpstart projects. Meanwhile, cloud, edge computing and 5G networking have co-evolved to boost the viability of IoT across a range of industries.
Companies offering IoT systems integration — the lineup ranges from regional MSPs to global systems integrators (SIs) — aim to keep pace with the changing state of technology and customer needs.
Consulting services begin with IoT basics
In today’s IoT market, service providers emphasize consulting clients before plunging into the technical issues of integration. This idea is to encourage customers to think more critically about their business needs and how IoT might fit into their plans.
“Our guidance for clients has been, ‘Let’s take a step back and take away some of these mad science experiments you have been working on and take it back to first principles,'” said Juan Orlandini, chief architect for the cloud and data center transformation division at Insight Enterprises, an IT services company based in Tempe, Ariz. “‘What is the business case you are trying to solve?'”
Orlandini advised customers to vigorously pursue an IoT initiative if it will help them make or save money. IoT experiments may still be appropriate, but customers must treat them as such. “Don’t overinvest. Do a proof of concept,” he said.
Michele Pelino, principal analyst at Forrester Research, also cited the need for customers and their integration partners to begin with the fundamentals.
“Starting with the problem is the first thing,” she said. “Solving the problem could be an IoT solution or it could be something else.”
Ultimately, the answer could be an “and” rather than an “or.” A customer may need IoT in addition to other technology offerings the integrator can provide, Pelino said.
Cloud, edge and the new integration
Cloud computing ranks among the top technologies now associated with IoT.
Myke Miller, consulting managing director at Deloitte Consulting and dean of the Deloitte Cloud Institute, cited the “natural linkage” between the cloud and the enormous volume of data generated by IoT sensors and devices.
“The cloud providers recognize IoT as a driving use case for their services,” he said. “I think that has been a motivator for them to add a significant level of sophistication and integration to their own products and services.”
The cloud shift, in turn, has fueled edge computing and the idea of “intelligence at the edge,” Miller added. Edge computing offerings such as AWS Outposts and Microsoft Azure Stack let organizations move computing workloads closer to where data originates. This avoids the latency and bandwidth issues of moving all IoT data to a central site, while pushing decision-making to the vicinity of the IoT sensors themselves, he said.
Those localities could include a car, an oil well or a factory, Orlandini noted. In the past, IoT sensors sent data to a central repository before AI and machine learning could process it to glean actionable insights.
Miller pointed to Amazon Lookout for Equipment, which became available in April, as an example of edge-IoT convergence. The AWS offering embeds edge capabilities into industrial gear for predictive maintenance. Such developments make it easier for end users to integrate IoT with business processes, he said.
The IoT integration potential of cloud and edge computing also changes the discussion between customers and service providers.
“It’s a much more strategic conversation,” Pelino said. “Not everything is going to the cloud, and not everything will be processed at the edge.”
There’s a change management piece of an IoT solution that SIs can bring to the table.
Michele PelinoPrincipal analyst, Forrester Research
This conversation flows from the initial consultative focus on the customer’s business problem. With the key use case in mind, organizations can address the salient questions, Pelino suggested: “What data and insight do I need at the location of the connected assets, and what should be sent back to the cloud for additional processing and advanced analytics?”
IoT off the shelf
Such high-level discussions are a decisive shift from the original crop of IoT integration services. “Back in the beginning, you had to be a really good designer or mechanical engineer,” Orlandini said. A customer could call on an integrator to create a custom enclosure to protect IoT gear from harsh industrial environments, he said.
Today, however, major hardware manufacturers offer boxes equipped with pre-integrated IT, Nvidia graphics cards, built-in 5G — all in an enclosure designed to withstand shocks and temperature extremes.
Service providers, meanwhile, offer their own take on out-of-the-box integration. Companies now provide IoT starter kits. Deloitte, for instance, offers turnkey IoT accelerators for predictive maintenance, asset performance management and asset tracking use cases.
The accelerators shorten the time required to build and deploy IoT applications, which have previously taken months to get off the ground, Miller said. Deloitte’s integrated IoT packages, which incorporate offerings from vendor partners such as AWS, Hewlett Packard Enterprise (HPE) and Predix, deliver a proof of concept (POC) in eight to 12 weeks, according to the company.
“We can very quickly iterate on IoT-based solutions,” Miller said.
Insight in 2020 launched an IoT foundation that it calls the Insight Connected Platform. The company’s accelerator aggregates a multitude of sensors and devices from Microsoft, Intel, Cisco, HPE and Dell Technologies, among other partners.
Forrester’s Pelino said such accelerators and pre-integrated offerings aim to help customers demonstrate IoT’s value sooner rather than later. “They don’t want to put years and years of resources, money and time into an effort,” she said.
While an accelerator can jumpstart an IoT project, the approach “may not address a lot of the comprehensive requirements” customers would like to see roll out in the future, Pelino added. A successful starter project, however, can pave the way for broader deployments.
The journey toward increased scale and enterprise-wide deployments provides another opportunity for IoT integration services. Systems integrators come into the frame when organizations seek to extend their POCs and limited IoT installations.
“Where you start is not where you end up,” Pelino said. “As you extend into new use cases, more business processes connect through the [IoT] solutions.”
In this context, integrators can help customers scale their IoT deployments, offering technology stacks built around their vertical expertise and partner ecosystems, she said. An IoT integrator’s advantage lies in providing customers with access to best-in-class technology and offering pre-integrated interfaces that link their various ecosystem partners or to legacy analytics and business system environments.
But technology isn’t the only consideration. An IoT project will often spark organizational change as it extends beyond a POC. A business might need to revamp how it performs particular processes while employees change the way they do their jobs.
“Those issues are more challenging than technical ones,” Pelino said. “There’s a change management piece of an IoT solution that SIs can bring to the table.”
Integrators can also help customers address skill set gaps, which, along with the rigors of organizational change, can deter scalable implementations, she noted.
Service providers seek to guide customers through the POC-to-production process. Insight, for example, uses three internals teams to execute projects. Insight’s Digital Innovation group assists customers with IoT ideation and POCs, Orlandini said. When the customer is ready to scale its initial IoT foothold, Insight’s cloud and data center transformation division determines what type of infrastructure and security the customer will need. Finally, the company’s connected workforce business handles the logistics of procuring, pre-integrating and shipping equipment for a customer’s IoT deployment. Insight’s supply chain optimization group supports the other divisions in providing IoT services.
However, as service providers pursue the future of IoT, they are sometimes called back to what Orlandini calls “first principles.” And the quirks of connected devices.
For example, when Insight took a restaurant customer’s POC into production, it found the system’s IoT base station wasn’t receiving signals from the temperature sensors installed inside the client’s commercial refrigerators. To remedy this, Insight had to custom design a bracket that mounted sensors inside the refrigerators where they could broadcast with the appropriate signal strength.
Orlandini said the fix was relatively simple once Insight took a close look and applied some fundamentals.
“The first principle is that radio does not travel well through metal objects,” Orlandini said. “You can’t ignore physics when dealing with the real world.”