How the IoT will transform manufacturing
The products of the future will impact not just factories, but all aspects of business.
Editor’s note: this post is an excerpt from “Enterprise IoT: Strategies and Best Practices for Connected Products and Services,” by Dirk Slama, Frank Puhlmann, Jim Morrish, and Rishi M. Bhatnagar.
In some cases, a distinction is made between the industrial IoT and the consumer IoT. But when we talk about “Enterprise IoT,” our focus is less about specific application domains and more about openness and integration maturity.
Here, we will take a closer look at some of the more industrial applications of Enterprise IoT, starting with a discussion about how IoT will transform manufacturing from the perspective of both product engineering and production technology.
Integrated production for integrated products
We believe that the IoT will have two main areas of impact on the current manufacturing landscape. The first concerns the organizational structure that is required to produce truly integrated IoT solutions. The IoT involves a clash between two worlds in which those in the machine camp and those in the Internet camp will be required to work together to create products that combine physical products with Internet-based application services. In an IoT world, many companies will discover that being just a manufacturing company or just an Internet company will no longer be sufficient; they will need to become both — or become subsumed in an ecosystem in which they play a smaller role.
For manufacturing companies, this means they will have to build capabilities in IoT service development and operation — in other words, the achievement of integrated production for integrated products. Many of these companies will find this challenging because it is not in their DNA. Nor is it just a question of developing additional IT skills (beyond the embedded skills most will likely already have). Value propositions will have to evolve, too, which will necessitate change in almost all parts of the organization, from engineering to sales right through to aftermarket services.
The second area where the IoT will have a significant impact on manufacturers is in the area of manufacturing technologies. As promoted by initiatives such as the German government’s Industry 4.0 strategy, connected manufacturing equipment, connected logistic chains, cyber-physical systems, and big-data-based analytics of production processes will help improve the way the physical parts of a connected IoT solution are produced. In a sense, this second area of impact can benefit from the first; what is an integrated product to one company — a machine component manufacturer, for example — is an advanced production technology to another (a manufacturer using the connected machine component in their assembly lines, for example).
Before we look at how new production technologies will help improve manufacturing processes in the future, let’s briefly recap what we know about the products of tomorrow. Ultimately, the nature of these new products will have an impact on all other processes, from design to manufacturing right through to aftermarket services.
Our assumption is that the products of the future will be connected and become part of what we call the Internet of Things. We are also assuming that products will have embedded computing capabilities, enabling local intelligence and digital services. These digital services can be applications or content. For example, a car app store might provide a new navigation application, and the application itself allows the purchase of additional maps.
The combination of physical product and connected backend service will have a sizeable impact on product design. First, it’s possible that the design of the physical products themselves will change. For example, a product’s embedded display and keys could be dropped in favor of a mobile app. This would constitute a significant redesign of the product’s physical components. Secondly, products will be increasingly reliant on remote services, often in the cloud. Building these kinds of related IT services is not usually part of the traditional product engineering process. It will require someone to oversee the design of both elements — the physical product and its associated backend software services or platform — and ensure that everything results in a nicely integrated product offering.
Finally, connected products will provide a rich source of product usage data, which will serve as input for all other stages of the value chain, from sales, marketing, and product design through to manufacturing and after-sales services.
Sales/marketing and new business models
New business models made possible by the emergence of the IoT will drive the future of product design. These business models will also have a significant impact on the sale and marketing of these products. Servitization involves transforming a company’s business model from one focused on selling physical products to one focused on services. For example, Rolls-Royce now earns roughly 50% of its revenue from services (by leasing jet engines to airlines on a “power-by-the-hour” basis, for example). This completely transforms the way in which products are sold and serviced.
However, it also means that sales teams will have to completely adjust their sales strategies. Incentive models based on upfront revenues will have to be revisited in favor of models that support recurring revenues, which allow for the stabilization of revenue forecasting.
Marketing teams also will be able to leverage detailed product usage data to drive marketing campaigns and define precise market segments. This direct link to the customer via the product can be of huge value for sales and marketing teams, making it easier for them to run targeted cross-selling and up-selling campaigns.
Another key driver is product customization. More and more, markets are demanding fully customized products. Ranging from custom-designed sneakers to cars built to customer specifications, this trend has two key implications. First, products are now being sold before they have been produced, and not the other way around. Secondly, this trend has a major impact on the manufacturing process itself; for example, “batch size 1” production is a basic requirement of custom manufacturing.
3D models are also playing an increasingly important role that transcends the traditional domain of product design. Modern 3D PLM systems have integrated CAD design data with bill of material (BOM) data and other information to better support end-to-end digital engineering. The 3D model becomes the master model for all product-related data.
3D data also support the simulation of entire assembly lines, helping to optimize manufacturing efficiency and minimize the risk of costly changes after the assembly line has been set up.
One of the key benefits promised by the IoT is that it will help link the virtual world with the physical world. 3D models are a very important type of virtual model. The use of sensors, lasers, and localization technologies has enabled the creation of links between the virtual 3D world and the physical world. For example, Airbus uses 3D data to emit laser projections over aircraft bodies in order to guide assembly line workers. Similarly, at the Hannover industrial trade fair in 2014, Siemens showcased a complete (physical) assembly line with an associated virtual model in their 3D factory simulation environment. Sensors on the moving parts of the assembly line send movement data back to the IT system, which then updates the position data in the 3D system in real time.
Augmented reality is another interesting area in which we are seeing convergence between 3D models and the physical world, especially in the context of training and quality assurance. For example, Airbus’s Mixed Reality Application (MiRA) allows shopfloor workers to access a 3D model using a specialized device consisting of a tablet PC with an integrated sensor pack. Leveraging location devices on the aircraft and on the tablet PC, MiRA can show a 3D model of the aircraft from the user’s perspective, “augmenting” it with additional production-related data. Airbus’s adoption of MiRA has allowed them to reduce the time needed to inspect the 60-80,000 brackets in their A380 fuselage from three weeks down to three days.
The factory of the future
One of the visions of Industry 4.0 is that it will enable the decoupling of production modules to support more flexible production. One potential way of achieving this is through the use of product memory. Products, semi-finished products, and even component parts will be equipped with an RFID chip or similar piece of technology that performs a product memory function. This product memory can be used to store product configuration data, work instructions, and work history. Instead of relying on a central MES system to manage all aspects of production, these intelligent products can tell the production modules themselves what needs to be done. This approach could be instrumental in paving the way for Cyber-Physical Systems (CPS), another key element of the factory of the future.
Improved “top floor to shop floor” integration is another important benefit promised by Industry 4.0. Concepts like manufacturing operations management (MOM) have emerged to help integrate and analyze data from different levels, including the machine, line, plant, and enterprise level. With IoT, additional data will be provided from the machine level directly.
Another interesting discussion relates to the integration that needs to take place one level down, i.e. at the bus level. For decades, industrial bus systems (EtherCAD, Modbus, Profibus, SERCOS, etc.) have been used for production automation, enabling communication with and control of industrial components, often via Programmable Logic Controller (PLCs). Most of these bus systems are highly proprietary because they are required to support extremely demanding real-time requirements — which is difficult to achieve using Internet Protocols (IPs). This, again, poses a problem for the overall vision promised by the IoT: the IP-enabled integration of devices of all shapes and sizes. So, it will be interesting to see if the efforts of the IEEE’s Time Sensitive Networking (TSN) task group succeed in establishing technologies for machine and robot control based on IP networking standards.
Other important technologies that could become relevant for the factory of the future include:
3D printing: 3d printing is set to become very important in the near future, primarily in the area of prototyping and the production of non-standard, low-volume parts.
Next-generation robots: Robots are already being used in many high-volume production lines today. In terms of how they will evolve, one interesting area is the ability of robots to work in dynamic environments and ensure safe collaboration with humans.
Intelligent power tools: Power tools such as those used for drilling, tightening, and measuring are becoming increasingly intelligent and connected. The tracking and tracing of these tools is an important IoT use case.
High-precision indoor localization: The tracking and tracing of moving equipment and products in a factory environment will be primarily achieved through the use of high-precision indoor localization technology.
IoT service operations
The ability to make the transition from manufacturer to service operator is essential to achieve success in an IoT world. This applies not just to the technical operation of the service, but also to the operation of a business organization capable of supporting strong customer relationships. The DriveNow car sharing service is a good example of this. Formed as a result of a joint venture between BMW and Sixt, the service successfully combines BMW’s car manufacturing expertise with Sixt´s expertise in running a considerably more service-oriented car rental operation. Another good example is the eCall service, an IoT in-car emergency call service that requires a call center capable of manually processing incoming distress calls from vehicles and/or vehicle drivers.
Remote software distribution is another area that offers a huge opportunity for many manufacturers, but also requires the provision and operation of a suitable infrastructure. A good case in point is the recent recall of 1.9 million vehicles by a large OEM due to problems with the on-board software. This OEM could have saved itself massive amounts of money if it had been able to distribute the required software update remotely. Smartphone platforms also provide a good insight into the challenges involved in running remote software updates on a very large scale. Although they are now much better at handling software updates than they were in the past, the situation is far from perfect and occasional problems still persist. In the case of in-car software, this would be unacceptable.
In an era of IoT-fueled “servitization,” aftermarket services are becoming increasingly important.
Remote condition monitoring (RCM) is one of a number of basic services that can have a fundamentally positive impact on customer service quality. The ability to access product status information in real time is invaluable for support services, especially because it makes for much more efficient root cause analysis and solution development. The challenge for many large manufacturers today is one of heterogeneity. A large manufacturer with thousands of product categories can easily have hundreds of different RCM solutions. The issue here is not so much the need for new and improved RCM for next-generation products; it’s about the implementation of efficient IT management solutions that are capable of managing this heterogeneity. This could be achieved by automating virtualization and improving secure connection management, for example.
The next step in the evolution of RCM is predictive maintenance. The use of sensors (for thermal imaging, vibration analysis, sonic and ultrasonic analysis, oil and liquid analysis, and emission analysis) allows the detection of problems before they even occur. For buyers of industrial components, predictive maintenance has the potential to significantly improve operational equipment efficiency (OEE). For end-consumer products, predictive maintenance is a great way of improving customer service and ensuring extra sales or commission: “You should replace your brakes within the next 5,000 kilometers. We can recommend a service station on your way to work.”
As mentioned earlier, product usage data will help with the identification of cross-selling and up-selling opportunities. When combined with the ability to sell additional digital services, the proposition becomes even more compelling. For example, the performance of many car engines today is controlled by software. We could have a scenario where a car manufacturer produces one version of an engine (the high-end version) and then uses configuration software to create a lower-performing version. The digital service in this case could be the option to temporarily upgrade engine performance for a weekend trip: “You have just programmed your navigation system for a drive to the country. Would you like to upgrade your engine performance for this trip?”
Naturally, this newly won customer intimacy will require solid security and reasonable data access policies in order to retain customer trust in the long term.
End-of-lifecycle data can be used for remanufacturing and recycling offers, or simply to make the customer an attractive product replacement proposals.
The boundary between IoT services and aftermarket services is not always clear. From our perspective, IoT services are part of the original value proposition. Take the eCall service, for example. In this case, the service is essentially the product that is being sold. Aftermarket services generally take the form of value-added services (which can also be IoT-based).
Other industrial applications
The industrial IoT presents many opportunities beyond those related purely to manufacturing. Some of these opportunities include:
Mobile equipment tracking: The tracking of industrial equipment and containers was one of the first application areas of telematics and M2M, and will evolve and contribute to value-added IoT solutions.
Nuclear physics research: One of the areas in which sensor technologies are most widely used is in nuclear physics research, where they are deployed to reconstruct digital images of nuclear collisions.
Energy: Energy is and will continue to be a large application domain for IoT.
And of course there are many other potential applications of the industrial IoT, from cross-energy management to mining to offshore drilling.
Public domain image on article and category pages via The Google Art Project on Wikimedia Commons.
Buy “Enterprise IoT: Strategies and Best Practices for Connected Products and Services,” by Dirk Slama, Frank Puhlmann, Jim Morrish, and Rishi M. Bhatnagar, now in early release.