What is the Industrial Internet of Things?
Thirty years ago, if you were the dispatch manager for a U.S. manufacturer of high-performance valves, your day-to-day schedule was a bear. Communication with partners took place by phone, fax, or mail. Staff manually logged inventory and checked out shipments for delivery. Vehicle fleets were maintained using visual inspections or regularly scheduled repairs. Drivers could be contacted by two-way radio if they were close enough, but in many cases, the only way to pass along information about a particular shipment or a new pickup location was to have the driver regularly check in … by pay phone. These inefficiencies existed up and down the supply chain, from the factory floor to customer facilities. It was primitive, prone to delays, and costly.
How times have changed. Modern industry now depends on advanced machinery, sophisticated software and communications technologies, and modern management techniques that emphasize lean inventories and integrated supply chains.
What is the Industrial Internet of Things?
But there's another trend taking place: Buildings, factory equipment, vehicles, warehouses, pallets, and workers themselves are increasingly connected to the network. Devices range from tiny environmental sensors to complex industrial robots. The networks they use may be wired or wireless. Some low-power devices will run on battery power or use power over Ethernet (PoE) connections. Others must be plugged in or are hard-wired to a building’s electrical system. Many of the devices have their own firmware and can be remotely updated or reprogrammed.
In other words, sensors, controllers, and specialized devices now exist at the edge of the network, assigned to unique IP or network address, gathering data and performing tasks that vastly inform and extend the capabilities of the enterprise. It’s the Industrial Internet of Things (IIoT), and it's transforming industries from manufacturing to transportation, construction, mining, and energy.
IIoT in practice
Consider the hypothetical manufacturer of high-performance valves. In the old days, managers could order new inventory only after calling down to the warehouse floor and asking someone to do a manual count. Orders for new parts were sent to suppliers by fax or express mail, with a handful of people required on each end to process the paperwork and accounting.
Today, the same company can harvest many more inventory insights, thanks to IIoT technologies such as radio-frequency identification (RFID) pallet tracking and warehouse bins that use cameras and weight sensors to calculate the number of parts on hand. This data is integrated with ERP and other software systems. When supplies of a certain part drop to a certain threshold, the system automatically places an order to the supplier for a refill. Humans seldom need to get involved with placing an order, which speeds up shipments across the supply chain and reduces the need for labor to handle what used to be a manual, high-touch process.
IIoT efficiencies are extended to the transportation of the finished valves. The days of truckers hanging out at phone booths and checking paper manifests are long gone. Instead, the company’s IT systems use optimization models to determine the most efficient use of fleets and track the products across the country with live trip data fed back to headquarters and updates passed to drivers on the fly via mobile apps or other devices connected to the network. RFID-enabled pallets are tracked as they leave the warehouse and are loaded onto trucks, and upon delivery are automatically registered at customers' receiving centers.
Onboard sensors that are part of electronic logging device (ELD) systems monitor speed and driving time and even how often individual drivers use their brakes, which helps conserve fuel, improve driver safety, and reduce idle resources. If a certain trucker executes a dangerous maneuver or is at the wheel for too long, the system will alert the driver and notify the dispatcher. Further, the ELDs meet federal monitoring requirements for drivers, replacing paper logs that drivers were once required to fill out every day.
IIoT also enables better fleet maintenance. Instead of bringing trucks into the shop for scheduled maintenance, onboard sensors can alert fleet managers to problems with specific vehicles. Trucks can be fixed or reassigned to easier routes, which cuts down on emergency repairs and delivery delays.
Similar transformations are taking place across all kinds of industrial companies. A German auto parts manufacturer uses IoT and other advanced technologies to manage its supply chain and fabrication facilities. A shipyard that builds submarines for the U.S. Navy relies on a network of RFID sensors, devices, and antennae to track tool usage and hundreds of thousands of parts. North American power utilities use high-resolution sensors to monitor the performance of their grids and proactively work with neighboring utilities and regional authorities to prevent blackouts and other problems. At a mine in Western Australia, massive driverless excavators equipped with cameras and hundreds of sensors have increased output and boosted safety.
According to IDC, global spending on the Internet of Things, including both consumer and industrial IoT, reached $737 billion in 2016, and will grow an estimated 15.6 percent per year through 2020. Industries making the largest investments in IoT in 2016 included manufacturing ($178 billion), transportation ($78 billion), and utilities ($69 billion). Growth has entered the hockey-stick phase: There are already more Internet-connected devices than people. Gartner forecasts more than 21 billion connected devices by 2020, but some sources, such as Beecham Research, point out that sky-high growth rates are unrealistic considering real-world capacity constraints—including insufficient resources to implement IoT.
While some companies are wary of technology buzzwords and vendor hype around disruptive technologies, IIoT is nevertheless a real trend that will impact many industries over the coming decades. Industrial vendors have already started to prepare for the anticipated wave of IIoT adoption, and are staking out vertical markets where they already have a presence or see a big opportunity. However, there are challenges facing companies that opt to implement IIoT in their operations, not to mention a great deal of uncertainty about where it will all lead.
While IIoT technologies have been available for more than a decade, the concept of IIoT being distinct from the IT and software revolutions that preceded it is relatively new. If IT connects people, data, and business systems, IIoT goes one step further, providing IP connections to a widely distributed network of sensors and machines that in turn increase productivity and drive new insights into the enterprise.
What's behind this trend? Certainly, recent technological innovations and adoption trends makes IIoT possible. PwC has identified the following IIoT enablers:
- Inexpensive new sensors
- Faster processors
- Cheaper storage
- Open platforms that can be integrated
- Vendors and partners encouraging adoption
- Greater functionality in core business systems
- Data analytics and machine learning
IIoT is more than a technology story. Many companies view IIoT not only as a way to reduce costs while speeding up production and delivery, but also as a platform for growth and a way to effectively compete, says Robert McCutcheon, U.S. industrial products leader for PwC.
"The IoT push is predominantly coming from the CEO and the board,” McCutcheon says. "It has become a common point of discussion in the boardroom around disruption and competition. A lot of this is a dialogue around, ‘What are our competitors doing from a technology perspective? What are customers doing from a technology perspective? Where might we be disruptive in that process, and what opportunities do we have to take advantage of and disrupt the market ourselves?'"
For IT managers, the wave of IIoT adoption is upending their traditional role. In companies that are implementing IIoT technologies, IT cannot be confined to the server room or desktop support. Rather, IT is increasingly seen as a crucial partner as devices and sensors go live on the shop floor and field sites.
Indeed, adoption of IIoT represents a redistribution of control over systems, processes, and knowledge centers. IT managers and staff—as well as their counterparts in manufacturing, operations, distribution, and supply chain management—must get out of their comfort zones to implement IIoT technologies and reap the benefits. This will be uncomfortable for some, particularly when it comes to letting go of power and sharing data outside of traditional silos.
IIoT efficiency gains and the talent shortage
Currently, industrial firms that are leveraging IIoT are often doing so to improve efficiency. Processes that can be streamlined and platforms that can be connected are obvious targets for IIoT integration. At a recent Georgia Tech workshop on IoT for manufacturing, Andrew Dugenske, director of the institute's Factory Information Systems Center, listed some areas that can potentially be addressed by IIoT:
- Asset and inventory management
- Capacity and performance monitoring
- Quality analytics
- Real-time alerts
- Safety and compliance
- Throughput monitoring
- Warehouse optimization
Dugenske believes that company culture has a big influence on whether or not a team will implement IoT. "Companies that appreciate the benefits of IoT and are willing to invest in new technology will realize a timely payback and competitive advantage," he says. "Companies that delay their implementation will be at a disadvantage that will be difficult to overcome."
PwC's McCutcheon concurs. "For some of the more progressive companies that find there is a key business imperative or strategic objective, it forces them down this path earlier than others," he says. "But it's becoming a bit more mainstream, and these conversations are becoming more commonplace. And as they become more mainstream, the progress you make is exponential. I think we're really at the beginning of the uptake."
However, as more and more companies turn to IIoT, they will increasingly run up against the challenge of finding qualified workers to design, implement, and maintain IIoT-based systems. "There will be a different skillset necessary to use and run and operate the technology-enabled environment that may be different than the old analog environment that manufacturers were working in," McCutcheon says.
Companies can work with outside stakeholders such as local governments and vocational training centers to develop a pipeline of technicians who can work in the factories of the future. But McCutcheon warns that firms will also need to overcome a perception gap. "The perception historically was that manufacturing wasn't necessarily the most high-tech environment to work in," McCutcheon says. "Manufacturers have to change that perception. Some of the most technologically advanced jobs we have are now in manufacturing. So it's also about attracting that talent, not just training."
IIoT data management challenges
IIoT systems generate staggering amounts of data. An oil rig outfitted with IIoT systems can produce up to 8 TB of data in a single day. Vehicles outfitted with IIoT sensors are even more prolific, ranging from a car (1 PB of operational data per day) to jet aircraft (with a single Boeing 737 engine generating 333 GB of data per minute). The amount of data will surge as more and more companies build out their IIoT infrastructures and more sensors come online.
What to do with all of the data? If the ultimate goal is to streamline operations, inform managers, and generate actionable insights, the data will need to be transmitted, processed, summarized, visualized, and stored. The IT requirements are formidable. A single IIoT-equipped building or vehicle may have thousands of sensors and devices sending data over wireless or wired network connections. Data may be processed with chips or dedicated hardware on the shop floor, remote campus, or vehicle control system. The data may also be converted or shared using APIs or custom software hooks, or put into storage.
When such systems are expanded to an entire company, including its core campus as well as satellite facilities, partner locations, remote sites, vehicle fleets, and employee devices, it becomes clear that IIoT has a big data dimension. This means big data approaches will be required to put the data to use, such as turning to cloud-based applications and storage.
There is a real risk of companies not fully leveraging the data that's available to them. A McKinsey report notes that just 1 percent of the data from an oil rig with 30,000 sensors will be examined. That’s because the sensors are primarily used for detecting problems, as opposed to optimization and predictive analytics. It's a wasted big data opportunity.
There is also the IIoT security angle to consider. This is uncharted territory for companies that are used to dealing with network security for data centers, office systems, and personal devices. In an IIoT environment, there will need to be an expanded focus on how to secure devices, data, communications, and processing power in locations that are far beyond the traditional reach of IT.
The very nature of IIoT hardware, software, and network connections has led to some unintended security consequences. Many low-cost devices such as remote cameras, routers, and even DVRs have limited security features (including hard-coded default passwords and telnet access) or are never patched by their owners. Once compromised, they become staging grounds for distributed denial-of-service (DDoS) attacks and the distribution of malware, among other problems. One nasty IIoT botnet, known as Mirai, has sowed Internet outages across the globe and continues to be a problem today.
As you might expect, an enterprisewide rollout of IIoT systems is extremely complex. Requirements include low-cost, robust sensors and other devices that can be integrated with existing systems. Managers also must determine how the devices will be powered and connected with data networks. Indeed, systems architectures may have to be modified to accommodate IIoT—and there may be significant integration challenges.
"Costs are plummeting for sensors, bandwidth, processing power, and communication capabilities, which is driving IoT," Georgia Tech’s Dugenske says. "But it does not address the lack of agreement on how entities communicate. A key question: What are the specific data structures and protocols that are used underneath? These huge cost drivers are typically glossed over but must be codified to be successful."
In theory, IIoT standards should ease integration, whether you are dealing with a generic controller or an industry-specific product. Companies can always turn to products that are built around existing standards, whether they be global IT standards (such as 802.11ac) or those that are expressly developed for industrial use, such as the IEEE Standard 1377-2012 for utility meters.
There is also a push to form alliances that are promoting the development of new standards. They include the Industrial Internet Consortium, which promotes common architectures, interoperability, and open standards for energy, healthcare, manufacturing, transportation, and smart cities.
Uncertainty can happen when competing standards arise, or a vendor sells IIoT devices that are incompatible with products or systems sold by another vendor in the same field. Take the Weightless-P open standard for low-powered wide area networks (LPWAN). Having a standard that can offer faster speeds over existing LPWAN technologies while using less battery power is crucial for IIoT vendors and customers alike. However, there are other LPWAN frameworks, including the proprietary LoRa standard and NB-IOT, a narrowband radio technology sanctioned by the 3GPP telecommunications standards group.
It’s not clear which standards will dominate, but it's a question that has big implications for IIoT, with an estimated 345 million connections (or about 26 percent of the total) taking place over LPWANs by 2020. Uncertainty creates potential integration headaches down the road, and encourages some industrial companies to take a "wait and see" approach before making a bet on a certain product line or technology standard.
The future of IIoT
Industry is still at the very early stages of a major wave of technological transformation driven by IIoT, big data, and other emerging technologies. This is not hype, but rather the inevitable consequences of advances in software, hardware, networking, and systems that are impacting nearly every sector of the global economy.
Some companies are already leveraging IIoT to increase productivity and develop new products and services. As prices for ever more advanced sensors, processing power, and data storage continue to fall, mainstream industrial companies—not to mention their smaller competitors and suppliers—will turn to IIoT to stay competitive and find new opportunities for growth.
Yet serious obstacles remain. Concerns about security, staffing, and interoperability may keep some industrial companies on the sidelines. Others may have real questions about benefits, costs, and returns on investment. Certain technologies are still in the early stages of development, held back by uncertainty over standards or technical and physical limitations (as with batteries).
There is also the challenge of bringing new technologies onto the shop floor or assembly line, into the vehicle maintenance bays, and onto remote sites. Modern industrial equipment may be sophisticated and powerful, but much of it is not yet connected to IP networks. Moreover, there are certain types of industrial machinery, often in little-known verticals or with very specialized uses, that may not be made IIoT-capable for many years to come. Replacing or retrofitting is possible—and indeed, as the pace of change accelerates, it will be an imperative. But for today's operations or IT manager, that day may seem very far away. In a 2015 survey of manufacturers, nearly one-half said they had no timeline or no plans to use data from smart sensors to improve manufacturing and operations.
Where some see reason for being cautious, others see a huge opportunity. Forward-thinking industrial companies, as well as startups in energy, transportation, and manufacturing, are making huge investments in IIoT and big data. Such companies want to position themselves for the future, and the possibility of using IIoT now to gain a significant competitive edge or develop greenfield products and services is irresistible.
Consider a world in which sensors, systems, data, and even robotics and artificial intelligence are ubiquitous. Assembly lines, supply chains, and remote sites will be largely run by autonomous equipment and algorithms, cutting down delays while increasing worker safety. Self-driving vehicles on the roads and in the air will be able to "talk" with the pallets they are transporting to better monitor their contents and coordinate deliveries. This is not the realm of science fiction, but rather a vision of what is possible with IIoT.