Almost every technological device today is equipped with smart sensors in one way or another. By Frederik Troester, head of connected devices and solutions, ASEAN, Bosch
Drawing on a study of 795 executives from a range of sectors in Asia Pacific, Europe and the Americas, Sreenivasa Chakravarti, head innovation and transformation group, manufacturing business unit of Tata Consultancy Services shares his insights on strategic decisions involving investing in the Internet of Things (IoT).
In times to come, the manufacturing space will be so interconnected that nothing will escape the eyes of those manning the operations – all in the name of increased efficiency and productivity. By Matej Kranjc, managing director of ASEAN and ANZ, National Instruments.
The Industrial Internet of Things (IIoT) promises a world of smarter, hyper-connected devices and infrastructure where manufacturing machines, transportation systems, and the electrical grid will be outfitted with embedded sensing, processing, control, and analysis capabilities. Once networked together, they’ll create a smart network of systems that shares data between devices across and across enterprises in the cloud.
To support the new capabilities of IIoT-enabled infrastructure, designers and end users alike need reliable, remote, and secure access to smart, leading edge devices. Network technologies must evolve to satisfy the requirements of these next-generation industrial systems and radically advance the way we operate our machines, electrical grids, and transportation systems.
Existing IT networks are defined by IEEE 802 standards, which specify requirements for different Ethernet layers and functions and ensure interoperability between devices. It defines standards and protocols for wired local area networks (WLAN), metropolitan area networks (MAN) and wireless networks; defines characteristics, operating procedures, protocols and services for networks that carry variable sized packets and specifies the development and handling of compatible devices and equipment.
Today’s hyper-connected world relies on IEEE 802 Network Standards to carry out daily tasks for both work and leisure. For example, computer, smartphone, e-book reader and gaming systems are just a few of the devices containing interfaces compliant with the suite of network interoperability standards developed by the IEEE 802 LAN/MAN Standards Committee (LMSC). Just imagine all the activities we undertake on these devices and what we couldn’t do without them.
Today, industrial suppliers, IT vendors, and silicon providers are collaborating within IEEE 802 and the recently formed AVnu Alliance to update standard Ethernet protocols and provide bounded, low-latency data transfer for time-critical data in IIoT applications, because the IIoT adds stricter requirements to its local networks for latency, determinism and bandwidth.
Time Sensitive Network
Industry consortiums are now working to address this challenge using a standard called TSN (Time Sensitive Network), a real-time Ethernet solution that will become part of the open platform communications architecture in the future. Its development was originally created for in-car applications that required fast real-time communication. Ethernet has been used for automotive applications since 2008, mostly as a method for diagnostics communication and data download.
Increasingly, the large bandwidth Ethernet provides compared to other automotive in-vehicle networking technologies makes it an obvious choice for emerging applications such as camera-vision systems and infotainment systems. There is also a huge potential for Ethernet to be used for backbone network communication throughout the vehicle. This could even include safety critical applications which enable piloted and autonomous driving – a major trend in the automotive industry today.
The IEEE has now adopted it as the IEEE802.1 standard. This development brings real-time Ethernet capability to chip level to enable cost-effective edge processing solutions to be added to instruments and devices to give them real time Ethernet capability. The AVnu Alliance, working with member companies such as Broadcom, Cisco, Intel, and NI, will drive the creation of an interoperable ecosystem through certification, similar to the way the Wi-Fi Alliance certifies products and devices to be compatible with the IEEE 802.11 standard.
With further development of the IEEE TSN related set of standards, it might be possible to mimic real-time communication performance and some of the safety-related features provided by the combination of both ARINC664 and SAE AS6802, which are applied in critical aerospace applications for example.
Ethernet may also have great potential for use in the energy industry. These applications will require 802.1 to work with other organisations to help create complete standards solutions for the issues they face, such as energy efficiency. Ethernet could be part of the ecosystem that provides better overall solutions.
The new TSN standard will provide numerous benefits, but notably it would improve bandwidth, security, interoperability and latency and synchronisation, over today’s standard and specialty Ethernet protocols.
Large data sets from advanced sensing applications such as machine vision, 3D scanning, and power analysis can put a strain on network bandwidth. Proprietary Ethernet derivatives commonly used for industrial control today are limited to 100 Mb of bandwidth and half-duplex communication. TSN will embrace standard Ethernet rates (1 Gb, 10 Gb, and 400 Gb versions are in the works) and support full-duplex communication.
Most of the lower-level field buses used today achieve security through air gap and obscurity. They are influenced by the automotive industry, for which air-gapped and closed CAN networks carry all the control and operational data. But recent security breaches have exposed the need to fully extend security into the critical lower levels of control infrastructure. TSN protects critical control traffic and incorporates top-tier IT security provisions. Segmentation, performance protection, and temporal composability can add multiple levels of defense to the security framework.
By using standard Ethernet components, TSN can integrate seamlessly with existing brownfield applications and standard IT traffic to improve ease of use. In addition, TSN inherits many features of existing Ethernet, such as HTTP interfaces and web services, which enable the remote diagnostics, visualisation, and repair features common in IIoT systems. As an added benefit, leveraging standard Ethernet chip sets drives component cost down by virtue of high-volume, commercial silicon, especially compared with specialty Ethernet variants that are centred on lower-volume, ASIC-based implementations.
Latency and Synchronization: TSN prioritizes the low-latency communication required for fast system response and closed-loop control applications. It can achieve deterministic transfer times on the order of tens of microseconds and time synchronisation between nodes down to tens of nanoseconds. To ensure reliable delivery of this time-critical traffic, TSN provides automated configurations for high-reliability data paths, where packets are duplicated and merged to provide lossless path redundancy.
As IIoT adoption continues, increased amounts of data and widely distributed networks will require new standards for sharing and transferring critical information. Just as an ambulance or fire engine receives priority among other traffic during an emergency, the TSN standard ensures that critical, time-sensitive data is delivered on time, over standard network infrastructure. Welcome to life in the fast lane with the IIoT.
Applying the Industrial Internet of Things (IIoT) to factories confers a host of operational benefits. By Advantech
Machinery and production automation systems need to be advanced enough to deliver high performance, and integrated enough to provide economical operation, yet must be based on mature products and methodologies offering sufficient reliability.
So why push for tightly integrated operational information and other advanced functionalities if individual pieces of machinery are running “good enough”?
The main reason is because harvesting, processing and analysing the correct data helps operational personnel make the best informed choices at their facilities, and enables management to optimise strategic plans throughout multiple locations. Simply put, advanced data analytics improves efficiency, reduces maintenance, and creates a safer work environment.
Fortunately in recent years, a number of device, communication, and software capabilities have developed in an interrelated manner—making it easier to extract and analyse manufacturing data.
When combined effectively, they can elevate “business as usual” manufacturing to “smart” manufacturing. In fact, in many ways automated manufacturing is already smarter than one might expect.
Machinery and process plants commonly employ control systems with many types of sensors. While the highly-touted Internet of Things (IoT) concept promises that one day all devices will become networked information providers, it turns out that the Industrial IoT (IIoT) already has countless sensors and other devices reporting data to higher level automation systems. Where the IoT is directed toward consumer convenience, the IIoT takes a laser focus on efficiency and safety.
Manufacturers such as Advantech offer a spectrum of hardware and software to facilitate gathering information from the lowest level sensor, or any machine, and routing it over a network to higher level automation, visualisation, and information systems. Automation controllers pre-process and package the raw information from sensors and other field devices. These devices are the “things” in the IIoT.
Industrial wired and wireless networks, working in conjunction with the Internet and cloud services, are the superhighway for moving information. This information moves from field controllers to human machine interfaces (HMIs) located on the plant floor and in control rooms, and from the HMIs to front office PCs and out into the mobile world of smartphones and tablets.
Smart manufacturing is a powerful trend, building on readily available hardware and software to take production operations to the next level of performance.
The Time To Implement The IIoT Is Now
Manufacturing businesses worldwide want to implement the IIoT to gather more data and improve operations. While these objectives have been present for many decades, it’s now much more feasible to implement the IIoT because of the technology advancements as expounded upon below.
Why Implement The IIoT Now?
- Most new devices offer smart connectivity
- Methods exist to enable traditional devices to become smart
- Controllers are proficient at handling smart data
- Standardised wired and wireless Ethernet networks are economical, powerful, and pervasive
- Specific industrial networking formats are common
- Open interfaces and numerous drivers are available to facilitate economic integration
- Communication methods are suitable for private and public clouds
- Mobile visualisation offers new ways to bring data to users
- Big data harvested from the IIoT can be more easily analysed
- Smart manufacturing adoption can occur in steps, with benefits realised along the way
More often than not, connectivity is the “killer app”. Consumer devices such as phones, watches, appliances, and even sneakers are commonly able to connect and interact with each other.
Similarly, industrial devices have moved from awkward and proprietary communication interfaces to standardised networks and protocols, often Ethernet-based. In today’s market, industrial manufacturing demands connectivity from most devices purchased. Even if the functionality is not immediately needed, it helps to future-proof investments.
For legacy devices using basic analogue and digital signals, or maybe simple serial communications, there are modules that can boost this equipment up on to contemporary networks and protocols. In this way, end users can choose an upgrade path that preserves their existing system, yet provides value by making their “dumb” devices smart, leading to intelligent machinery.
Connecting Islands To The Mainland
Many production plants consist of “islands of automation”. Often, there are many automated skids or systems with minimal interaction among them, even though taken as a whole they form a production line. Sometimes these systems have been assembled and grown over a long period of time.
What they have in common, though, is that each island is operated by one or more controllers. Industrial controllers have more than enough power to perform some data processing, but may not share common communication protocols.
Fortunately, there are many flavours of “gateways” or “bridges” available. These can take the form of dedicated configurable devices, or PCs running various drivers and communication software. These gateways can translate pertinent information from existing systems into a suitable format for higher level integration.
When disparate controllers and the systems they control are capable of being connected, some huge informational advances can be achieved. Such systems can be interconnected to supervisory alarming and historian systems, consolidating key information from a whole production line into a few effective displays or reports.
For many operations, when subsystems are integrated in this way, it is possible to achieve a transfer of upstream and downstream information and improve the production flow. Or, when production goes down it is possible to use the integrated information to identify and eliminate the root cause, promoting overall equipment effectiveness (OEE) tracking.
These are just a few of the benefits of a connected factory. As Jamie Carter puts it, “In the wider economy, the IIoT is critical in reducing unplanned downtime of production facilities and plants.”
Moving Information To The Next Level
Assuming that technical and cost barriers are overcome for gathering information in a smart factory, what are the next steps? The first is typically to make the information visible to operators and managers so that they can make informed decisions.
This used to mean tabular lists or printouts of numbers, but information presented in this manner is difficult for people to process. That is why so many variants of graphical display software and HMI packages have been developed.
Earlier generation HMIs used to just reside locally to their associated factory processes. Today’s HMIs use networking, the Internet, and public or private cloud services to extend their reach to wherever users are. Instead of just a single machine, production line, or factory being coordinated—it is now possible to manage multiple factories across the world in a more organised manner.
The Internet and cloud services are ideal for publishing smart manufacturing information to laptops, tablets, and smartphones, putting the information directly in user’s hands. Many visualisation software packages have features specifically adapted to mobile device operation. It has become especially prevalent and useful for mobile devices to present a streamlined “dashboard” view which shows only the most important information in an easy-to-read format.
End user expectations from HMI packages have soared, due to consumer familiarity with high performance home computers, phones, and tablets. The graphics must be informative and must also look good and easy to use. HMIs that take advantage of multi-touch swipe and zoom gestures position themselves that much close to the everyday user.
Browser-based products like Advantech’s WebAccess are available that offer a familiar user experience, are easily extendable to all types of devices, and are able to publish the information conveniently over the Internet.
Harvesting Big Data
But the smart factory is about much more than just dishing out pretty graphics. At the factory level, the proper flow of status and command information is crucial for manufacturing execution systems (MES) that strive to track and record the production of finished goods. At an even higher level, data is required for enterprise resource planning (ERP) and business logistics systems to be effective.
A real opportunity exists when all of the big data can be harvested from many IIoT sources, and then effectively analysed to reveal inefficiencies that can be overcome or trends that can be intelligently re-vectored.
Gathering enough of the right information can enable users to make discoveries that would be otherwise impossible. Besides just improved throughput, benefits can be found in material costs, energy efficiencies, labour costs, maintenance costs, and the cost of adverse quality.
Keep in mind that implementing smart manufacturing is not an all-or-nothing proposition. If fact, adopting smart technologies and methods can (and often should be) carried out in steps. This reduces the initial cost, and allows an organisation to determine which pieces of the smart factory yield the most benefit for their situation.
The time to implement the IIoT is now, and here are the specific components which make up a typical IIoT implementation in a manufacturing plant.
IIoT Building Blocks
Data flowing through the smart factory can be imagined as a pyramid structure as shown graphically in Figure 1, and as detailed in Figure 2.
Another good reference is ISA-95, which defines industrial automation interface concepts from the lowest (Level 0) to the highest (Level 4) level in terms of both functionality and immediacy. If “Level 0” is considered to be the actual physical process, then the smart manufacturing foundation begins at “Level 1” and consists of the sensors and field devices.
Examples of IIoT building blocks:
- Smart sensors
- Network-capable I/O
- Controllers–PLCs, PACs, DDCs, Proprietary
- Network switches, media converters, routers, security
- Visualisation, fixed location
- Visualisation, mobile
- Business strategy systems
Traditional sensors were historically hardwired and offered only a single basic process signal, but today’s smart sensors are networked and provide additional process signals and device diagnostics. They can maintain on-board calibration data, and technicians can interact with these sensors remotely. Think of a flow transmitter that also provides temperature and pressure information, and can alarm when the data readings are suspect.
More advanced analysers can simultaneously provide multiple-sensed variables for complex parameters. Barcode readers and RFID tags are key ways to establish material tracking. Many other types of smart sensors and field devices are available, all capable of providing data to higher level systems.
The Highest Levels Of Smart Manufacturing
HMIs are “Level 2” systems that facilitate detailed plant operations. They can be PC-based running software, or a more dedicated hardware type. Plant networks supply HMIs with the information they need, either directly from field devices, or more commonly through I/O and controllers.
These HMIs can be flexibly located in main control rooms, on machines, in maintenance and management locations, or elsewhere. More recently, it has become common to configure consumer-grade or industrial-grade tablets as HMIs and troubleshooting stations that can be carried around the factory.
One of the real game changers in HMI space over the past decade is the emergence of browser-based products. No longer are users tied to specialised hardware, or difficult software installations. Just as PCs and Ethernet successfully leveraged commercial technology into the industrial arena, browser-based products prospered by offering much of the same end user experience as traditional software, but at a lower price point and requiring near-zero configuration on the end user’s device.
These products are capable of providing an HMI interface anywhere within a facility, on all types of mobile devices, and throughout the world via the Internet. Not only that, but they can offer advanced features such as integration with Excel, Google Maps, and video streams.
Comprehensive Smart Manufacturing Solution
Residing above HMIs are “Level 3” MES and “Level 4” ERP systems. These software-based systems typically run on servers located at a given production plant, or even far away in a corporate office. Software systems at each progressively higher level are typically less “real-time” than at lower levels. While MES and ERP systems are a subject of their own, they both require close integration with lower level sensor and control systems in order to be effective.
A comprehensive smart manufacturing solution built on an IIoT foundation is necessary to power operations and business management. These IIoT building blocks can be combined to create real-word applications to deliver specific benefits, as shown in the following example.
Any time there are multiple steps in a process, it is critical to identify which steps are the limiting throughput factor. Similarly, if there is a failure, then operators need information to point them to the root cause. Smart manufacturing will harvest all of the production key performance indicators, and use them to identify bottlenecks that can be improved, and will also facilitate troubleshooting.
At the highest level, data provided via smart manufacturing allows business operators to track, direct and optimise their raw material usage and productive output. Uptime and downtime can be analysed, and inefficiencies identified and wiped out. Without the data provided by smart manufacturing systems, none of this is possible.
Putting Your Data To Work
For today’s factory, superficial good looks aren’t enough to prove that things are running at their best. Instead, additional improvement opportunities must be actively sought to create a smart factory. One way to do this revolves around obtaining more operational data and putting it to work. Any process of improvement is based on quantitative analysis of measurements, and fortunately the IIoT opens up a whole new world of quantifiable data.
Connectivity is no longer a unique luxury, as it has instead become a baseline requirement. Intelligent machinery leads to a connected factory, which in turn provides the platform for smart manufacturing. Businesses everywhere want to leverage the IIoT in the most expedient way possible, and fortunately the technology is available now to make this happen.
The building blocks are smart devices, methods for making legacy equipment smarter, robust networking, and a wide variety of software—all of which are readily available to build into new facilities or integrate into existing operations. The widespread availability and ease-of-use of these enabling technologies allows end users to focus less on how to harvest the data, and concentrate more on improving operations.