From trend to trade: How automation evolves out of necessity

What was unthinkable in the past has become commonplace in modern warehouses. Robots and scalable automation are now must-haves for thriving during market fluctuations.

Warehouse automation has been around a lot longer in the U.S. than people may assume. Though today we think of autonomous robots and software-driven machines operating continuously in human-free environments, automation looked different (and meant something different) in past decades. Its modern-day roots grew during the Technological Revolution, also known as the Second Industrial Revolution, which ushered in the 20th century. The birth of mass production called for new technologies and systems to make manufacturing and distribution scalable and more efficient. However, space and labor were key factors that kept the U.S. in manual mode much longer than other countries.

“Warehouse automation in Europe and Japan was probably 10 years ahead of the U.S. in terms of pick up because labor was more expensive in those countries and space was more restricted. Here in the U.S., companies had a lot of land and affordable labor,” said Markus Schmidt, president of Swisslog Americas, a global warehouse automation provider. “But that equation has, of course, turned over the years.”

COVID certainly changed the nature of warehouse operations since market demand shifted rapidly and e-commerce skyrocketed. Plus, after COVID, getting people to work in distribution centers was hard, so there was a drastic labor shortage amidst increasing consumer demand.

Before COVID, warehouse automation focused on conveyance and sortation for goods-to-person applications. Over the past five years, especially during COVID, interest in autonomous mobile robots (AMRs) increased to reduce labor dependencies, increase efficiency, and afford rapid scaling in the event of sudden market changes.

CarryPick is a modular storage and order picking system that can be designed for any space and moved or modified easily. Image courtesy of Swisslog.

Take online grocery shopping, for instance. The dawn of mobile apps and real-time information exchange made it easy for consumers to order groceries online and have them delivered or ready for pickup at the store. While demand was manageable, personal shoppers walked the aisles and pulled items from shelves to fulfill individual orders. However, during the pandemic, online grocery sales increased suddenly, and safety protocols restrained manual fulfillment.

In response, new technology and small distribution centers emerged and picked grocery items efficiently and accurately with less labor. Such automation was unthinkable years ago because there was no need for it. Now, most advanced distribution centers have goods-to-person robotic solutions for item storage and picking.

“There are a lot of new companies with innovation entering the space, and there are a lot of established players. The industry overall grows about 20% year over year,” said Schmidt.

Balancing storage and throughput

Back in the day, when labor was plentiful, the buzzword wasn’t “automation” — it was “throughput.” In the early 20th century, transportation advancements made it possible to deliver goods faster to a broader customer base. The focus was on making more products quicker and getting them into customers’ hands.

But mass production came to a screeching halt during the Great Depression, which some claim was partially due to the overproduction of goods. As the economy started looking up, the Digital Revolution, or the Third Industrial Revolution, spawned around 1947 and paved the way for computers, advanced networking technology, and eventually automated storage and retrieval systems (AS/RS).

Yet as markets fluctuated, warehouses still felt the pangs of storing too much inventory that didn’t move. So, throughput became a balancing act, and companies started valuing efficiency, flexibility, and scalability — both up and down.

Until recently, if a warehouse needed more throughout, it would add more storage capacity and labor. In today’s automated systems, throughput and storage aren’t strictly proportional.

For example, Swisslog’s CarryPick storage and order picking system includes robots that carry mobile racks of goods, and warehouses can scale robots and racks independently.

“You can add more robots to get more throughput. Or you can add more racks to get more inventory,” said Schmidt. “That flexibility is especially valuable for online businesses and all warehouses during peak retail seasons.”

Modern automated warehouses are flexible and can independently scale inventory or throughput. Image courtesy of Swisslog.

Adding or expanding automation

Once an automation system is in place, it’s relatively simple to scale it to meet demand. However, how does a warehouse justify expanding or adding new automation, especially for goods with special storage or handling requirements?

Continuing with the grocery example, in an e-grocery warehouse, there are typically three major product categories that can be differentiated by temperature zone: ambient for dry goods, chilled for fresh produce, and freezing for frozen foods. Before COVID, frozen foods had relatively lower volumes than chilled and dry groceries. So, the first attempts to automate grocery order fulfillment focused on chilled and dry foods. However, as frozen food demand has increased, there’s an opportunity to expand automation for this product category.

But in the case of frozen food, the automation technology must withstand freezing temperatures to operate properly. Plus, the frozen food section of the warehouse and all environments where frozen goods are stocked, stored, picked, and transported must maintain strict temperature requirements — not to mention any pickup holding areas or delivery vehicles.

With all the extra considerations, warehouses need to be sure that they have enough volume to justify automating new product lines with different, strict requirements. Or they need to consider an adaptable and scalable solution that can adjust throughput seasonally or in response to market changes.

One solution is to set up one or more micro fulfillment centers to address varying demand and product requirements. Micro fulfillment centers are small, automated distribution centers located close to the consumer with limited inventory and throughput to serve a geographical area.

The big challenge with micro fulfillment centers is sizing. If they are too small, the return on investment may be disappointing. But if they are too big, there may be too much distance between the distribution center and the consumer. Warehouses need to find that sweet spot that balances size and distance. For instance, a superstore that sells many different goods constantly may need one large distribution center with different temperature zones to fulfill orders for 3-5 nearby stores. Other companies may find it practical to set up multiple micro fulfillment centers, each tailored for specific product requirements.

Micro fulfillment centers can also benefit online-only retailers. Companies that want to expand to a different geographical area can build a small center to serve online customers only with no investment in a physical store. Brick-and-mortar companies are already experimenting with this concept by diverting online orders to online-only micro fulfillment centers and dodging store fulfillment completely.

Design, data, and decisions

Systems such as Swisslog’s AutoStore take the guesswork out of warehouse automation. AutoStore is a standardized product with robots, specifically shaped plastic bins, and an aluminum grid on which the robots drive. Warehouses can select the size of the grid; how many bins, robots, and pick stations to include; where to locate the pick stations; and then Swisslog does the heavy lifting.

“We look at a specific client application, figure out how much inventory they need, how much throughput they need, and then we design a solution specifically for that requirement using the AutoStore technology,” said Schmidt.

Product data drives AutoStore system design to ensure the solution meets all inventory, throughput, and operational requirements. Image courtesy of Swisslog.

Automation companies such as Swisslog rely on customer data to analyze and understand the customer’s specific requirements and capabilities to tailor solutions. That means warehouses must provide lots of detailed product and operational data, especially when turning over manual processes, such as picking, to robots.

In highly automated distribution centers, the trend is to have robots automate the picking process. However, not everything is pickable by automation.

“I would say in today’s world, probably 70% is pickable, and 30% is not,” said Schmidt. “You need to make sure you get that percentage correct, and then you also need to design a way of dealing with products that are not automatically pickable because those you still need to pick conventionally.”

Though robots take the limelight when building a fully automated or hybrid system, advanced software is the glue that connects and runs the warehouse. Integrating technology to operate as one cohesive system is complicated and gets interesting when robots, automated processes, and humans need to interact. Software needs to guide both technology and people to do the right things at the right times while monitoring equipment, inventory, and locations and communicating with other software, such as customer-facing apps that send notifications on order status.

Software orchestrates all the moving parts, synchronizes data, and helps make decisions. If a robot on the floor sees an obstruction in its path, whether nearby or far away, it will either need to stop, go around the obstruction, or navigate a different pathway. Plus, software is needed to instruct the robot to make local decisions or communicate with a human operator who can tell the robot what to do. Some decisions must be made locally in milliseconds, while others can be programmed more strategically in advance.

AutoStore saves space and time with compact robots and vertically stacked bins organized on an aluminum grid. Image courtesy of Swisslog.

An industry ripe for bright minds

Today, the buzzword is Industry 4.0, which translates Fourth Industrial Revolution into modern-day digital speak. With advanced technology and the adoption of cloud computing, the world is more intricately connected, which can make it quite brittle, as COVID revealed. Though history may repeat itself, the current trajectory is an upswing for automation.

As warehouse automation seeks to decrease dependence on manual labor, there’s still plenty of room for engineers to build lucrative careers. Remember, what was unthinkable a few decades ago is now commonplace. More than ever, companies value innovation and creativity and want to stock bright minds to stay behind the economic cue ball. No one really knows what the future holds, but scalability, flexibility, and robots have already shifted from trend to trade.

“In schools and universities, there’s a lot of teaching around automation and manufacturing. But I think we as a society have squeezed out the last little bit,” said Schmidt. “In warehouse automation, where everything was conventional, we are just now entering the phase where real interesting automation takes place, and it’s going to continue for a while.”

Swisslog
swisslog.com