2019 Technology Roundtable: Are we ready for what’s next?

Hard to hear, but the daunting list of challenges facing logistics and supply chain professionals hasn’t changed much over the past two years. Freight rates continue to rise, capacity remains tight, global trade is more complex than ever, e-commerce growth isn’t slowing down, and warehouse/DC operations can barely keep pace with fulfillment as they continue to struggle with finding, training and retaining a workforce.

Continually asked to do more with less in an environment that’s speeding up and becoming more sophisticated, logistics and supply chain managers now—more than ever—need technology platforms that support good coordination and collaboration among stakeholders. But what to apply and when continues to baffle even the savviest practitioners.

This month we’re joined by four top technology analysts to discuss how some of the hottest software and technologies are helping logistics and supply chain management professionals streamline their operations to meet the pressing demands of digital commerce and manage through the tightest labor market in a generation.

Joining us this year are Daniel Merchan, research affiliate at MIT’s Center for Transportation and Logistics; Dwight Klappich, research vice president at Gartner; Joe Vernon, senior manager of North America supply chain analytics at Capgemini; and Norm Saenz, managing director at supply chain consultancy St. Onge Co.

Artificial intelligence (AI): Yes, it’s already happening

Logistics Management (LM): How would you best define artificial intelligence (AI)?

Dan Merchan: AI can be broadly defined as the ability of a machine to perform cognitive tasks such as identifying patterns, learning relationships, reasoning and representing knowledge. The ability to identify patterns, for instance, makes AI-powered technologies better at predicting outcomes. AI algorithms, including those based on machine learning (ML), speech recognition, natural language processing technologies, are already powering systems that predict movies we might like or the fastest route from home to work based on current traffic conditions.

AI can also boost innovation, insight and decision-making, capture efficiencies and drive revenue growth. Yet, AI is not a recent technological advancement. In fact, this field has been undergoing developments for more than 60 years now. For several decades, the cost of collecting massive amounts of data and processing large-scale AI models was prohibitive. Over the past 10 years or so, the convergence of enhanced AI algorithms, the availability of massive amounts of data to train models, and an ever-increasing computing power have energized the adoption of AI across industries. Put simply, making better predictions has become significantly cheaper.

LM: What role is AI already playing in logistics and supply chain management?

Merchan: Applications of AI in the supply chain space are expanding rapidly. Demand forecasting problems, for instance, are a natural fit for the prediction improvements associated with AI, and more specifically with ML systems.

The German e-retailer Otto has reported replacing its classic forecasting toolset with a deep learning model trained with large and multi-dimensional datasets to more accurately anticipate customer orders and effectively preposition inventory to reduce delivery lead-times. Other known applications include predicting travel times and operational disruptions. Also, AI-enabled technologies are driving new levels of warehouse automation.

LM: Are there any areas where you see AI playing a more important role in logistics and supply chain management?

Merchan: Demand forecasting is perhaps the low-hanging fruit of AI in logistics and supply chain management. The convergence of AI and robotics to drive warehouse and vehicle automation is another promising area for development. More broadly, I see two important roles for AI in the logistics and supply chain space.

First, it can expand the capabilities of existing decision support systems. Decisions in supply chain operations have traditionally been informed by optimization-based prescriptive approaches or simulation-based descriptive tools. Expanding these decision-support systems with enhanced predictive capabilities associated with ML can lead to more robust decision-making processes. For instance, improving vessel travel time predictions, which depend on a variety of factors from port scheduling to weather across multiple geographies, can better inform network planning and inventory allocation strategies.

Second, I see it unveiling new strategic insights. Let’s take as an example vehicle routing in last-mile delivery operations. Optimizing the sequence of customers to visit is a well-studied problem in logistics and routing solutions are widely available. However, the increasing availability of transactional, geo-location and telemetric data can help reveal new insights about customer delivery preferences, driving patterns and urban context conditions that could significantly redefine how routes get planned.

Unsupervised machine learning methods can help derive classification of customers, products or services, based on diverse and multi-dimensional attributes, to ultimately target strategies to specific customer types and operational conditions.

LM: How do you see AI evolving in logistics management? What are its limits?

Merchan: I think AI-enabled systems can play an important role in improving logistics and supply chain operations in the near future. Predictive analytics will continue to permeate through the different decision-making levels. Technological capabilities will also expand. For instance, we might see Siri-like virtual assistants for drivers and warehouse operators.

Many companies are already deploying armies of AI-powered robots for automated inventory storage and retrieval, eliminating some of the most physically taxing or monotonous tasks for operators. As robots become cheaper, safer and faster at learning new tasks, we might also see AI-trained robots assisting picking and other high-precision tasks.

Yet, logistics and supply chain managers must also separate the hype from the reality, and be aware of common misconceptions about AI and its limitations. While data availability has greatly expanded, data quality is still a major obstacle for rapid deployment of AI-based solutions.

LM: How can companies begin the AI journey?

Merchan: Start simple and put together a team with complementary skills. That team needs to bring together machine-learning scientists, data analysts, developers and logistics and supply chain domain experts. They need to start by developing a prototype solution to address a business problem in which prediction plays an important role.

You’ll need to make sure the team engages with decision makers to understand expectations and data communication preferences. Initial tech investment should be relatively low because many of the most popular ML tools are open source. Assess the value of the solution and understand scalability challenges such as data quality and IT integration. Becoming a data-driven supply chain organization is more about the people and the culture and less about technology itself.

Workforce Management: Attract, engage, retain

LM: Some have described the labor situation in logistics as “crisis level.” Do you believe we’ve hit crisis mode?

Dwight Klappich: I hate to be an alarmist and say that it’s a crisis, but all indications are that it’s very close to becoming one. If you look at the established economies and exclude immigration, most are in negative population growth, meaning that their populations are declining. To quote one of my German contacts: “No kid in Germany grows up wanting to drive a fork truck for the rest of their life.” You could say the same thing in North America, most of Western Europe and many countries in Asia.

The U.S. Bureau of Labor Statistics projected that employment in logistics and transportation would grow 7% and 6%, respectively, between 2016 and 2026, meaning that demand will continue to grow while supply declines, so competition for operational workers will become even more constrained over the next decade.

LM: What are the key drivers for the current labor situation?

Klappich: Where in the past labor shortages might be cyclical, up one year down another, all indications are that labor is and will continue to be a constrained, demanding and increasingly fickle asset. Recently, another Gartner client told the story of panicking upon learning that Amazon planned to open a warehouse nearby. The client worried about how the company would be able to compete with Amazon for people and how it could keep the e-commerce giant from cherry-picking their best people.

Bottom line: With more demand for labor, fewer people willing to do the work and more competition for the labor that’s there, companies now have to compete aggressively for every good employee. However, finding, on-boarding, keeping and growing the workforce of the future will be harder than ever before.

LM: How are you seeing labor management systems (LMS) evolve to keep pace with this tight labor environment?

Klappich: Labor and workforce management applications are certainly not new, but for most of their existence the focus has been on reporting labor performance retroactively. This remains meaningful, but leading vendors are beginning to focus on how their software can address employee engagement.

Can the LMS provide the Fitbit-like data that encourages employee engagement? Are there tools that support coaching more than punishing? The answer is yes and several vendors are focusing more in these areas, which bodes well for operational labor in companies that adopt these tools.

We have also seen a significant increase in interest in forward looking labor planning and forecasting capabilities. Like with inventory, if an asset is constrained then planning becomes increasingly important—which is now true in labor as well. Short term, within a day or shift, a company wants better visibility into the work and the demands this places on labor by area. The manger also wants to see how work progresses throughout the day to make sure that labor is most effectively deployed to support the needs of the business.

Longer term, companies are increasingly looking for tools to help them predict their labor requirements out in time. Today, labor forecasting—if done at all—is done using rudimentary spreadsheets lacking any sophistication. At best, companies can say they use X amount of labor in the same period last year, but will bump that up by 10% next year.

Finally, to encourage a new labor force like working mothers or college students, companies have to rethink their shift patterns and are considering tools to help them be more flexible and proactive in allowing workers to set their own schedules.

LM: As technology evolves, so too does an organization’s approach to labor management. How are successful organizations combining technology and improved standards to help retain employees inside logistics operations?

Klappich: Regrettably, operational labor, often referred to as blue-collar labor, has long been treated as a fungible asset. Economists define “fungibility” as “the property of a good or a commodity whose individual units are essentially interchangeable.” Simply put, one unit has no greater value than another unit, so the units can be substitutable.

Too many organizations have historically placed little value on operational labor when capacity was great and demand was low. Pay was low, pressure to perform was high and fear of losing one’s job was used to “keep workers in line.” Tools, if used, were aimed at measuring granular-level performance with an eye toward punitive actions.

For example, Sally performed well, so she gets to stay; however, Joe performed poorly, so he’s put on an improvement plan, while Sam had a few bad weeks, so he’s fired. According to this belief, this helped drive performance improvements through productivity gains because—as a fungible asset—people saw themselves as replaceable so they would theoretically work harder to keep their jobs.

Managing with a big stick might have worked in the past, but it won’t today. Too many companies still consider operational labor, an easily replaceable commodity, so they’re treated and paid as such. These will be the companies that struggle the most in the years to come. Today, it’s not simply about measuring people with an eye toward punitive action; it’s about building an environment where people want to work for your company.

LM: What does an ideal workforce management scenario look like/act like when both software and forward managerial thinking are applied to nurture and retain a workforce?

Dwight Klappich: Today, smart companies recognize the importance of engaging with employees to find and keep good talent. Consequently, operational labor management is evolving and technologies to support it are evolving as well. Companies need to take a holistic approach to capturing the hearts and minds of operational labor. To find and keep and advance their best people, companies need to expand their view and manage, attract, engage and retain the best employee group possible.

Managing operational labor has to evolve away from a “punitive” mindset to an “encourage” philosophy. Things like training, education and coaching become more important than simply penalizing bad behavior. The onus for process improvement shifts from the worker to management, where supervisors are measured on employee engagement and improvement over time, not just point-in-time metrics.

Warehouse/DC Operations: Think before you leap

LM: Based on your time inside facilities over the past year, how would you best define the current environment inside today’s warehouse and DC/fulfillment operations considering the myriad pressures—e-com boom, labor shortage, tighter delivery windows?

Norm Saenz: We’re seeing many of our clients rushing to design and implement major changes inside the warehouse for the very reasons mentioned. Growing volumes, increased e-commerce, and lack of labor and space are forcing many into taking swift action. This is the reality of today’s warehouse and DC environment, but we caution that a more detailed analysis of your specific needs is always recommended to make quality decisions.

LM: Is there a way to roll up the biggest challenges facing warehouse/DC operations considering this environment?

Saenz: The biggest challenge we’ve seen this past year is companies not allowing enough time to design the optimum warehouse solution. In many cases, operators are not affording themselves—or the design and consulting company—enough time to properly plan. We typically require eight to 12 weeks for a detailed facility master plan, but we’re increasingly being asked to squeeze the planning process into four weeks to six weeks.

In addition, clients continue to have a lack of quality data to evaluate and use to select the proper storage and material handling equipment. That situation leads to having poor WMS functionality employed in the operation. With a lack of dimensional and weight data at the item level, even a quality WMS is held hostage in releasing its full functionality.

LM: Are you seeing any predominant technology trends being applied to help tackle these significant challenges?

Saenz: I hate to raise doubt regarding the use of new technology available today. However, the warehouse and DC industry at large needs to collect product data and allow sufficient time/analysis to ensure the right technology is selected for enhancing an operation. Many larger organizations do have the required planning horizon, data and ability to evaluate automated design solutions, but there are others that need to get more prepared for the journey to automation.

We have created our own innovation center at St. Onge to stay up to date on the latest technologies, and are even developing our own tools, such as virtual reality simulators, to evaluate advanced design solutions. The use of autonomous vehicles, robotics and Internet of Things (IoT) technologies are increasingly part of our alternative evaluations.

Moreover, companies that historically only used traditional rack and lift trucks are now considering conveyors, control systems, put-walls, pick-to-light, voice and other technologies. These tools are quickly becoming the norm in the industry.

LM: What seems to be working best in terms of meeting the most common challenges?

Saenz: The common issues prevalent in today’s warehouse and DC operations—besides the lack of data and systems—include running out of space, equipment (locations/storage), and labor. The first approach to dealing with these challenges is designing efficient process, material flow, layout configuration and slotting.

Most of these items don’t require any capital; yet can yield improved use of space, equipment and labor. For many common challenges, basic industrial engineering can be the answer. At the very least, you should start with these types of low- or no-cost solutions before investing in technology.

LM: Are there any tried and true pieces of equipment or software that continue to be overlooked?

Saenz: Smaller companies that are quickly growing are stuck using carts and paper-based systems. For them, moving from carts and paper-based to basic conveyor-based designs and WMS/RF technologies is like going into “warp-speed.” These companies are likely not ready for—or likely do not need—robotics and the latest technologies on the market.

Instead, proven technology and best practices can push these organizations to the next step in their warehouse evolution. They can more than double their efficiency by simply using proper storage equipment, layout/material flow, WMS/RF technology, labels and signage, forward pick area with conveyor/control technology, packaging systems, and shipping systems.

LM: So many warehouse/DC professionals are feeling overwhelmed. What practical advice do you have for operations managers who are helping their facilities undertake a transformation in 2019?

Saenz: Give yourself time to get ready and start collecting the data required for the digital transformation. Depending on your current level of sophistication, evaluate the use of advanced technologies and systems, and make sure they have a solid ROI. You also need to make sure that any new technology is the right solution for your company and growth plans.

GTM/Blockchain: New levels of collaboration needed

LM: Are the new complexities of global trade pushing more shippers to invest in and implement global trade management (GTM) software?

Joe Vernon: To put things in perspective, let’s consider the most recent numbers available. The 2015 GTM software market was valued at roughly $670.5 million. According to market research firm Technavio, the projected market value for GTM software will balloon to $1.193 billion in 2020. This growth matches the general trend across supply chain solutions to embrace technologies that improve the immediate communication between vendors and suppliers and remove overhead and paperwork from day-to day-processes.

LM: What functionality of GTM will prove most helpful in managing today’s uncertainties?

Vernon: I’d have to say that improved robotic process automation (RPA) has certainly helped streamline data input inefficiencies and reduce rote tasks. RPA gives employees additional time to complete higher value-add tasks like analyzing historical data or reaching out to customers. I believe that GTM will provide similar efficiencies by automating monotonous tasks related to regulatory 
compliance and logistics.

GTM is intended to offload and cut through some of the complexity and historically time-consuming aspects of international trade and commerce. International vendors and suppliers sit in different time zones and speak different languages and abide by varying rules and regulations. The right GTM attempts to manage these variables and allows employees to redirect their attention to the issues that will typically require the human touch.

LM: What advice would you have for those global shippers now leaning in the direction of applying GTM to their operations? What steps should they take?

Vernon: It’s difficult sometimes with emerging technology to pick the right one, and companies will tend to migrate toward a brand name—only to later realize the solution functions and capabilities aren’t aligned well to their operational needs or industry.

That said, my first piece of advice is to map out current operational procedures and pain points and be sure to gather feedback from your employees. You and the team will want to organize prioritized lists based on which manual and problematic processes require the most time. This should help highlight the areas where software and automation can add the greatest value to your business.

You’ll then want to meet with multiple providers and score them against your list and criteria for success. And if you can match with a vendor that’s using a software-as-a-service (SaaS) model, then all the better because this way you’ll minimize your overall upfront investment and can switch out that vendor more easily when the next best thing comes along.

LM: There’s so much chatter about blockchain, and now we’re seeing a few marque shippers (Walmart, Nestle, Unilever, Tyson Foods, Starbucks) getting involved in a food safety blockchain. What stages are these early adopters in, and what would you say is the top driver behind their investment?

Vernon: Last October we surveyed roughly 450 supply chain business executives that were exploring blockchain to get a better understanding of rate of adoption. What we found was the majority (97%) of participants are in an early experimental or proof-of-concept stage and only 3% had a working solution in production. We also asked executives to list the key drivers that led them to blockchain experimentation.

The top five answers were cost savings (89%); enhancing traceability (81%); enhancing transparency (79%); increasing revenues (57%); and reducing risks (50%). In my own opinion, it’s traceability, provenance and food safety that will continue to drive wider implementation of blockchain in the near term.

One of the companies you mentioned, Nestle, has seen promising results from blockchain experimentation. They’re using the technology to trace the lifecycle of product ingredients that can help ensure food quality and safety. In food recall scenarios, blockchain helps determine where the food contamination began—preventing substantial losses and more accurately targeting where to alert consumers and reducing the need to recall such large batches of product.

I thought it was interesting to see cost savings rank as the top implementation driver. I think that’s a reflection of the increased pressure on supply chains to deliver more and more capability like last-mile and full end-to-end visibility while holding down costs. For many of these executives, blockchain is potentially the light at the end of the tunnel.

LM: What do you see as the biggest potential benefits of blockchain, especially from a global logistics perspective?

Vernon: Blockchain has the potential to enable what I would call “the seamless self-driving economy.” That’s a commercial model of contracts and terms, near 100% visibility of all movements and transactions, and machine driven negotiating and decision making for the procurement, movement and fulfillment of goods and services.

If you can get these terms of engagement and visibility to span geographies, then trade and international commerce may move toward self-regulation. At a minimum, it will be a platform for transparent, secured and trusted ledger systems in society. Watch the adoption rate of “smart contracting” within a blockchain as an indicator of unlocking the full future potential.

LM: What are the biggest barriers to blockchain implementation?

Vernon: It’s an evolving technology and capability and there will be some companies who may never have a reason to join a blockchain. But how do you know when and if it’s right for you? If it does look like something you could use, then you’re still faced with two main issues—timing and risk of adoption.

First, it’s difficult for leaders to know when the right time is to invest and to what extent. The technology providers are working diligently to create blockchain platforms, business cases and easy to implement solutions; however, these are the early days of adoption and that raises the risk profile for any technology.

Second, in order for blockchain to be successful, companies will have to collaborate in a new and more expansive way. It’s one thing to be excited about a new way of doing things, but quite another to get everyone on the same page and actually taking that leap of faith to trust a new system to take control of a big part of their business. Supply chain partners will need to commit to implementing and working out the kinks.

LM: For global shippers putting a toe in the water on blockchain, any advice?

Vernon: There’s a definite trend among companies learning about blockchain to execute a series of small-scale pilots or proof of concept initiatives. These provide a low-risk, low-cost experiment that allows vendors and companies to get to know the technology and each other and take an iterative approach to understand how the blockchain might impact and elevate their business.

Vendors are offering SaaS blockchain solutions that can help companies keep upfront costs and investments down and allow for payment based on use. Find a simple use case and partner with a ‘lighthouse customer” to see what you can achieve quickly and cheaply. And do not fear failure. You may not see the immediate value, but you will have educated yourself and will be more prepared when the right business case or market imperative compels you to make blockchain a permanent part of your organization.