Supply chain teams and networks are facing mounting pressures caused by rising and shifting customer expectations for product variety and service levels. In the past, when supply chain networks faced additional fulfillment needs and customer service requirements, they responded by ramping up hiring (also known as "throwing more people at the problem"). But a constantly tightening labor market—from the frontline to the C-suite—has made this strategy unsustainable. It shouldn't be surprising then that a growing number of supply chain network leaders have hit the tipping point and are turning to technology solutions and technology-based business practices to reduce the pressure.
The number of emerging technologies being pitched to the supply chain market have certainly proliferated. High on the list of mentions include blockchain, artificial intelligence (particularly machine learning), autonomous vehicles and robots (including drones), augmented reality, Internet of Things, additive manufacturing, big data and predictive analytics, mobile devices, asset-sharing systems, cloud-based technology, and control towers.
But which of these are truly ready to be rolled out broadly and which are still in the development stage, best left to the early adopters? Based on interviews with industry experts, this article identifies the emerging technologies that are ready to meet your supply chain challenges and others that, while generating tremendous interest, are not yet ready for broad adoption.
What does "emerging" mean?
Just about every large consulting company, analyst firm, or multinational association seems to have developed a list of the top emerging technologies based on a variety of criteria. But what constitutes an "emerging" technology? One of the best definitions of emerging technology comes from the World Economic Forum's (WEC) "Top Ten Emerging Technologies 2018" report.1 The report defines emerging technologies as those that are in the early stages of development and are generating excitement and increased investment. While they are not yet in wide use, emerging technologies will likely provide significant benefits to societies and economies within the next three to five years. Ideally, there should be more than one company developing the technology. Finally, according to the WEC report, emerging technologies will disrupt or alter industries and "established ways of doing things."
We believe that few technologies are truly new in capability or concept. Many emerging technologies are add-on or replacement tools, providing incremental changes as part of age-old continuous improvement processes. Even when they are new, they often ride up and down and then up again on the hype curve (sometimes not making the final rise at all or doing so in a different form). But which of the many new technologies have the potential to ride out the hype curve and meet your supply chain challenges in the next three to five years?
The Industrial Internet of Things
The emerging technology holding the greatest promise is the Industrial Internet of Things (IIoT) and the complementary concept of Industry 4.0. At its core, IIoT is about acquiring a greater and wider range of more timely data using process sensors (for temperature, humidity, pressure, vibration, location, or others) and automatic-identification transactions (by using vision or voice-based systems such as bar codes, RFID, or others), then processing that data to support better, faster supply chain resource allocations.
These types of data acquisition technologies are not all new. What is new and emerging are the tools, such as control towers and edge computing, for channeling, filtering, and analyzing this flood of data to provide visibility of demand and asset status.
IIoT shows tremendous promise for providing the real-time information needed to make better decisions about whether you can meet (or promise to meet) demand and about how to allocate your resources quickly enough to meet that demand. It enables the original promise of digital transformation to provide better distributed decision making. The challenge, however, is integrating all that data using data models that are consistent for all the tools that are involved in the production and distribution handoffs among suppliers, operations, customers, and partners.
When it comes to implementing IIoT, it makes sense to invest in several, prioritized pilot projects across your operations. Developing an all-encompassing, centralized set of infrastructure and rules before implementation is likely a strategic mistake because it is too big of an ask and takes too much time to plan and accomplish. While quick prototyping will lead to a lot of smaller failed experiments, it will also produce unseen big wins that central planning can't envision or produce. That's because smaller teams that are closer to the problems can more quickly define and implement prototype solutions that can then be integrated into more centralized plans and spread throughout the organization's supply chains.
Autonomous and collaborative robotics
Second up on the near-term emerging technology payoff list are robots and other automated solutions that augment human actions and mental tasks rather than trying to just replace or mimic people. The increased interest in robotics is being driven by the fact that labor has become more expensive and harder to acquire. Many general managers of manufacturing and distribution facilities are now focusing on how they can have humans do more of what they do best while reducing the time they spend on what they do worst.
Robots and other devices are already available that do a far more accurate job than any human at specific, well-defined, repeatable tasks, like reading bar codes or strings of numbers and letters; conducting high-speed visual inspections; palletizing like-sized boxes; and transporting bins, cases, and pallets from point A to point B. Plus they can do it tirelessly. People are not good at performing extremely monotonous tasks. They will get bored and lose attention, particularly when being driven at high speeds for long periods of time. Furthermore, they are not good at copying or recording information. Similarly, if your people are spending their time walking around (for example, moving from one pick location in a distribution center to another), then you are wasting a valuable resource.
Robots are also easier and quicker to retrain. Retraining can happen in mass with an update, even while they are working at their current tasks. As Dwight Klappich, vice president at the analyst group Gartner put it, "Teach one robot, and you can teach them all, instantly."
In spite of these advantages, there are still some jobs in the manufacturing and distribution world that are more cost effectively performed by humans. People are still the best investment for picking a wide variety of sizes and shapes of productsthat are mixed together in a container and then placing them in another container or getting them ready for shipment. They are also good at seeing unanticipated patterns and making connections between observations.
In short, companies should look for opportunities for robots and other devices to assist people with the moving, inspecting, and data recording tasks, while letting picking and decision making continue to be performed by people.
Artificial intelligence and machine learning
Just as robots and automated material handling equipment can augment human actions, artificial intelligence (AI) and machine learning can augment human decision making. Once again, the labor is divided between machines and humans, with each performing the tasks that they do best. The machine can tirelessly and accurately look at vast amounts of data and report on connections it has found through a variety of pattern-matching, neural-network training, or other techniques. Then a trained human can determine whether that observation makes sense and whether an action needs to be taken, plus whether that new learning should be applied automatically going forward.
AI should currently be treated as a highly capable scanner of data that brings situations to a human's attention, rather than as an autonomous decision maker. That's because humans have not yet defined the world well enough for AI to understand situational context and when rules should or should not be applied. AI does not currently have the wisdom to act on its own except in the most well-defined situations where few exceptions exist that could cause big problems. For now, focus on AI as an enhancer, not a replacement for people.
Successfully deploying AI and machine learning is not easy and requires significant capabilities and skills. You can't just buy an AI solution today, plug it in like a utility, and expect it to spit out brilliant solutions. You need skilled human "navigators" to guide and evaluate the machines and to determine how to use the insights and predictions that the technology produces. It is currently very difficult to find skilled people who understand how to structure and sift through data and how to configure AI. People with the right skills are currently in very high demand. Most firms' near-term best option, therefore, may be to contract with a services firm at a higher expense until it is financially attractive and realistic to acquire dedicated internal resources.
Dense, flexible 3D flow systems
Goods-to-person material handling systems have done more than just emerge, they are becoming standard practice at high-volume, high-variety facilities with short response time. The storage and retrieval concepts are not new, but the configurations, tools, and uses are still evolving and are finding their way into smaller, more urban facilities closer to end customers. For goods-to-person applications, shuttle-based or bot-based sortation systems that can operate flexibly throughout three-dimensional spaces, rather than just within aisles, are an important part of that evolution.
These systems allow companies to use their resources more flexibly while still remaining focused on the core principle of providing efficient access to a wide variety of products within ever-denser storage systems. Companies no longer have to commit to a fixed number of picking resources within an aisle, bay, or level of a storage rack or shelving space. This flexibility is crucial because the speed of change in customer demand requires less dependence on a ramp-up of labor. People are still a very capable, very flexible part of this future. But they can now be combined with flexible automation systems that can shift resources to an area without getting in each other's way. Automated and human resources can be intelligently deployed to areas where current demand shifts have dictated that more resources are needed—whether that be for the next few minutes or for the next few hours.
These shifts in demand and resources cannot be effectively planned in advance through central planning. Instead they must be sensed and responded to using pre-planned rules and human evaluation, augmented by the suggestions provided by information systems. Emerging technology to support that flexibility starts in the form of sensory and data-gathering capabilities that instantly provide data on demand and resource status and trends. Then, systems like control towers gather that wide variety of data to determine ways to redeploy resources throughout facilities or networks. Siegfried Zwing, president of redPILOT, a control tower for warehousing and distribution, describes these systems as being like car navigation systems. "Like car navigation systems, they advise you on how to deal with current conditions and suggested paths forward for the resources you have, except now you want your system to meet order requirements or cost constraints," he said.
Asset-sharing systems
The sharing economy is typically looked upon as a growing trend rather than as an emerging technology, but these new systems for resource sharing depend greatly upon many emerging technologies and techniques. IIoT, big data analytics, mobile devices, and cloud-based systems are all part of what make asset-sharing systems scalable. For this reason, we are including asset-sharing systems as an emerging technology.
While consumer-facing systems such as Uber and Lyft have received the most attention, asset sharing is also being applied in the industrial realm. There are many startups and long-time existing transportation companies that are trying to apply the asset-sharing model to the movement of freight and parcels. Likewise, the shared use of manufacturing capacity, distribution capacity, and even services capacity is also possible but will require huge amounts of data about demand, coupled with huge amounts of data about resource use and availability.
In supply chain, it is currently more about vehicle fleets and space, but every product and service is on the table, and you need to keep variations of the words "asset sharing" on your radar. Your focus should be on evaluating whether you should own supply chain assets that vary greatly in their utilization or whether you should rent or share those assets. Evaluate the idle time that some of these assets have and look for opportunities to utilize shared resources to meet the peaks and valleys of demand. You may also be able to share your own resources with others if the investment has already been made.
A reality check
Unlike the previous emerging technologies, there are several others that the sources interviewed for this article felt were not yet ready for prime time. While many of these technologies do hold great promise, there was a general feeling that they were being prematurely hyped as ready for widespread adoption. Chief among those were blockchain, additive manufacturing, and augmented reality.
Blockchain. Blockchain has huge potential for enabling fast, secure, and visible transactions among partners in supply chain networks. Part of the blockchain promise is the ability of trusted systems to automatically share and trigger approvals of transactions and payments based on predefined conditions. Electronically enabling transactions that automatically trigger contractual actions or payments can lead to great efficiencies, as anyone familiar with the evolution and capabilities of electronic data interchange (EDI) understands.
In spite of all its potential, blockchain is still currently low in its evolution and emergence because there is not yet a clear path for efficient scalability across a wide range of trading partners. Instead, it currently tends to rely on one or two large influential trading partners dictating the use of blockchain within a limited, predefined chain of partners. Its goal, however, is to connect together a much more complicated ecosystem (made up of a collection of businesses, roles, and industries) than has been linked together before. Signing up to participate in such a system requires a high level of trust in a set of previously unlinked trading partners. Creating that level of trust requires significant effort. It can be difficult enough at times to get two departments in the same company to share information or two companies to agree on data transfer and sharing policies. Blockchain is attempting to extend that level of trust and cooperation among a much larger set of players.
Given this level of complexity, most companies should let the early adopters of blockchain work the bugs out of the system (particularly for scalability of transaction volumes) before they even think about piloting the technology, let alone trying to pursue larger, critical projects. For now, it is best to rely on more traditional means using proven enterprise capabilities for sharing information with supply chain network partners.
Additive Manufacturing. Next up on the list of technologies that are not quite ready for scale is additive manufacturing. Additive manufacturing shows great potential for the on-demand production of critical replacement parts or for product customization. But while additive manufacturing currently works well for prototyping and visual proof of concept, it is not yet ready to be used for higher volume production.
However, that does not mean that you shouldn't be conducting pilot projects with the technology. Your company should still be investigating how additive manufacturing can be used to serve customers (both internally and externally) in new ways. But you should likely not be buying your own systems until you are ready to scale up their use, as the tools and level of expertise involved are quickly evolving. Instead work closely with companies that either specialize in additive manufacturing or companies with additive manufacturing capabilities that supplement their already proven capabilities in contract manufacturing.
Augmented Reality. While other technologies that augment human capabilities, such as robotics and AI, are proving to be capable of producing value today, current applications of augmented reality (AR) have not found a great deal of traction. In spite of great implementation examples in entertainment, tourism, and games, AR has not been widely adopted in industrial settings. The primary exception is in the areas of maintenance and manufacturing tasks, where AR is helping workers to learn and better perform a wide range of tasks. AR is enabling better evaluation and guidance during tasks rather than a reliance on post-process quality control checks after mistakes are made.
The adoption ramp-up of AR could accelerate if applications are developed that enhance both humans' capability to perceive the status of a task or orderand needed actions and if that information could be fed to workers in a way that was helpful rather than distracting.
Companies that are not already doing so should explore use cases for AR in maintenance and manufacturing that involve replacing or enhancing the use of manuals. They should also look at whether employee training could be enhanced by learning to do a task first in an AR or virtual reality (VR) environment, similar to flight simulators. Be aware that the future may bring a wider range of applications when AR and VR no longer require costly custom applications development.
Focus on emerging needs, not technologies
It is easy to get excited by emerging technologies and their promise to enhance your supply chain network. Or to become terrified about their potential to enhance your competitors' supply chains first before you can catch up. Or to worry about their potential to cripple your own network if you don't implement them in the right way.
But it is important to remember that it's never about technology first. The desire to pursue the "gee whiz" aspect of technology is tempting, but it can be potentially dangerous for your career, as it can lead to costly company "learning experiences." It is much better to first clearly identify the needs of supply chain network partners in terms of competitive direction and customer satisfaction, and then try to find the solutions or emerging technologies that support those needs. As Gerald Hofer, CEO of Knapp AG, a supplier of highly automated supply chain solutions advises, "Start with the prioritized needs of the final customer, then back up through the chain, evaluating processes and the ranges of flexibility needed. Only then is it okay to discuss how tech can support the processes."
As Hofer recommended, where possible, focus first on the final customer (or at least your customer's customer). What delights them? What do they wish they could do better? Then work your way backward through your supply chain and start diving into how these new technologies could be applied to your business.
Consider, for example, the grocery industry. "A quickly growing and evolving customer demand in our industry is the number of our customers who wants us to shop for their groceries," says Kevin Condon, senior director of engineering and supply chain network strategy for the grocery store chain Kroger.
Instead of starting with an emerging technology first, the grocery industry started with this customer need or desire. The industry has been trying to fulfill that desire for many generations in a low-tech way and has been through many cycles of successes and failures, particularly around the time of the dot-com bust in the early 2000s. The industry is now seeking to fulfill these demands by using many of the emerging technologies discussed earlier in this article.
One of those enabling technologies is high-density storage systems that use armies of bots or shuttles in a three-dimensional grid to store and retrieve goods based on demand patterns identified by machine learning systems and guided by distribution professionals. Although automated storage and retrieval systems have been around for years, the addition of machine learning has made these systems more flexible. The bots or shuttles can now be sent to the same aisles to meet constantly changing throughput levels while intelligently being routed to stay out of each other's way. They then deliver the goods to human pickers to do what they do best with their eyes and hands rather than wasting time walking.
These enhanced storage and retrieval systems are also being applied in new emerging ways. The startup company Takeoff Technologies has seen value in placing these systems nearer to customers, and even within stores, rather than placing them only in regional distribution centers. "Dedicating in-store space to goods-to-person systems using dense, flexible storage provides a 10x advantage in costs over having a lot of people perform picking throughout a traditional store," said José Vicente Aguerrevere and Max Pedro, the company's co-founders.
Whether storage systems are in-store, nearby, or in regional distribution centers, companies will need to pair them with increased visibility to the status of the order and of supply. "No matter how well you manage facilities for outgoing customer demand, a critical factor is the visibility that a supplier is going to show up on time with the right quantity of the right product," Condon said. He sees a strong interest, both in grocery and in other industries, for emerging visibility systems, like the control towers mentioned earlier, to direct resources efficiently across distribution networks based on demand patterns. However, he warns that for these systems to work, they must have good, clean, meaningful, and timelydata.
The path forward
What do you need to do to prepare your supply chain networks for emerging technologies? First, get better at asking about and knowing the needs, capabilities, and metrics of your entire supply chain network, which should include suppliers, customers, and partners. Second, get better at understanding the capabilities of the emerging technologies and matching those capabilities to your networks' needs. Third, quickly develop, prioritize, justify, and allocate resources to a range of pilot projects that use these emerging technologies as a form of innovative research and development. Small pilot projects will allow you to dip your toes into the water and build trust in the technology's capability and security. Having this foundation will help you scale up in response to supply chain network opportunities and challenges. In 2020 and beyond, if you aren't behaving like a technology-based solutions incubator, you better be really good at acquiring that expertise from the marketplace.
Finally, and most importantly, you need to understand that the days of developing and implementing technology just to make your own internal operations work faster and cheaper are gone. In the future, technology solutions will need to provide shared value among a network of partners, likely in a cloud-based ecosystem that provides strong capabilities for data gathering, filtering, analysis, advice, evaluation, communication, resource allocation, and measurement.
Those "working and playing well together" skills that you learned in elementary school now need to become one of the core capabilities in your toolbox so that you can take full advantage of the emerging technologies that will address your supply chain challenges.
Notes:
1."Top 10 Emerging Technologies 2018," World Economic Forum (WEF), www3.weforum.org/docs/Top10_Emerging_Technologies_report_2018.pdfÂ
