Robotized and automated warehouse systems

A robot may not injure a human being or, through inaction, allow a human being to come to harm. Three Laws of Robotics - Isaac Asimov

🧠 Learning objectives

• Discuss the rationale for automation.
• Identify and classify different types of warehouse automation technology.
• Analyze the adequacy of automation technology for different scenarios.

Why automation?

• Potential savings:
  • Space;
  • Labor cost (high employee turnover);
  • Operational costs (e.g., heating and lighting).
• 24/7 availability.
• Scalability and throughput flexibility.

Caveat: Automation of storage and order picking requires considerable scale and a long-term vision, as the investments can be earned back only in the medium and longer term Azadeh et al. (2019)

“The advantages of automation are mainly savings in space, savings on labor costs, 24/7 availability (it is not always easy to find unskilled personnel willing to do warehouse work), and savings on other operational costs, such as heating and lighting. Furthermore, robotic technologies provide scalability and throughput flexibility, which is essential in e-commerce environments where the demand variability is high. Automation of storage and order picking requires considerable scale and a long-term vision, as the investments can be earned back only in the medium and longer term.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 939).

The ‘Real’ Cost of High Warehouse Turnover Rates (kanelogistics.com) The latest figures from the Bureau of Labor Statistics put annual warehouse turnover rates at 43%. Losing that percentage of a 100-associate warehouse team would cost, at the very least, $365,500 every year. And that includes ONLY the easily calculated costs for things like:

• Departure costs – exit interviews, severance pay and increased unemployment taxes
• Supplementary costs for overtime or temp staffing
• Additional advertising, recruitment fees, and screening costs
• Drug testing, background checks and possible relocation costs
• Orientation training, certifications, uniforms and informational literature

Dark warehouses

Figure 1: Dark warehouses are fully automated warehouses that operate without human intervention, using advanced technologies such as robotics and guidance software for accurate order fulfillment (example by Addverb).

https://addverb.com/

What are Dark Warehouses and how do they work?

Dark Warehouse are designed to be fully automated, using advanced technologies such as robotics and guiding software for accurate order fulfilment. These technologies allow the warehouse to operate without human intervention, reducing costs and increasing efficiency. There may be minimal human intervention in lights-out warehouses when it comes to more complex processes, such as goods receipt and the packaging and labelling phase, but most of the installation is automated and remains dark.

Advantages of Dark Warehouses-

1. Increased Efficiency With the help of advanced technologies to perform tasks quickly and accurately, dark warehouses are designed to be highly efficient. Because everything is automated, there is no need for human workers to move products around the warehouse, reducing the time it takes to complete tasks and increasing overall efficiency.

2. Reduced Labour cost Dark warehouse operation needs little or no human intervention. This reduces labour costs, as there is no need to hire workers to move inventory.

3. Inventory management The dark warehouse layout powered by best-in-class software suite prioritises storage and self-pickup options, which greatly optimises SKU management, particularly for retailers that manage large SKUs.

4. Improved Safety Unlike the traditional Warehouse set up, dark warehouse operates with minimal human intervention, reduces the risk of accidents and injuries.

5. Dark Warehouse Solutions Behind setting up a fully automated warehouse – lies the combinations of highly efficient solutions. These includes AMRs for material movement, dense racking for storage and software to manage the inventory flow within the warehouse. In Addverb, we combine technical expertise and our hardware solutions moving towards dark warehouse.

Setting up a Dark warehouse? This is what you need – Automated Storage and Retrieval Systems for improved space utilisation – Arranging items and using vertical storage systems helps warehouses store more goods. Item grouping and shelving are crucial in warehouses and distribution centres to operate efficiently. Sorting robots for High speed parcel sortation – The sorting robots move goods and materials to different locations within the warehouse. They operate independent of human control; they need human help only to tend to the placement of objects on the inbound end or to monitor the reject lane and other output locations. Mobile robots for quick Material movement – These robots can move boxes and pallets in a busy warehouse all by themselves, thanks to wireless commands from a computer system and laser-guided eyes that help them find their way.

Looking for Warehouse Automation Solutions, we are here to help you. Mail us at automate@addverb.com or visit our website: - https://addverb.com/

The vendor description above illustrates the commercial “dark warehouse” concept. The following scholarly passage establishes the same concept in the warehouse-systems literature:

“Pioneered by Witron, combining multiple technologies has led to the advent of completely automated warehouses, particularly in the store-based retail industry (mostly grocery). Based on the authors’ experience, in Western Europe alone, about 40 fully automated warehouses are in operation, and many are under development. Although these warehouses are large, they are much smaller (and supposedly more cost-efficient) than their conventional, manual counterparts.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 918).

Material flow in a typical automated warehouse

Figure 2: Material flow of an automated retail warehouse with typical storage and handling¹ systems. No manual handling is involved, except for the technicians responsible for maintaining the system (Azadeh et al. 2019).

1. In such an automated retail warehouse, selected suppliers unload their own trucks and feed the preannounced single–stock-keeping unit (SKU) pallets to a check-in conveyor (Step 1).
2. The pallets are then stored in an AS/R system (Step 2).
3. When a certain product is requested, the pallet is offloaded and automatically destacked (Step 3).
4. The loose cases are then often put on trays to ease manipulation and are stored in a miniload AS/R or in an AVS/R system (Step 4).
5. When the store order arrives, the cases are retrieved and sequenced (Step 5), and
6. mixed-case palletizers build the pallets or roll cages in a store-specific sequence that allows rapid shelving in the store (Step 6).
7. These roll cages then wait in an order consolidation buffer (OCB), usually an AS/R system (Step 7), until the departure truck arrives, after which they are retrieved and loaded in the sequence determined by the stop sequence in the truck route.

Apart from the (many) technicians needed to keep the system alive, no manual handling is involved.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 918).

1. OCB: Order Consolidation Buffer; AVS/R: Autonomous Vehicle Storage and Retrieval

Classification of automated picking systems

• Automated picking systems
  • Crane/Automated Forklift
    • Single/Double-Deep
      • AS/R system
    • Multi-Deep
      • Push-back rack
      • Conveyor-based
      • Satellite-based
  • Carousels and dispensers
    • Horizontal/Vertical carousel
    • Vertical lift model
    • A-Frame
  • Shuttle (AVS/R)
    • Aisle-based
      • Single/Double-deep storage
        • Horizontal
        • Vertical
        • Diagonal
      • Multi-deep storage
        • Multi-deep shuttle/transfer car
    • Grid-based
      • Dynamic storage (puzzle-based)
        • GridPick
        • Live-Cube
        • GridFlow
        • GridSort
      • Static storage
        • Robotic compact S/R system
  • AGV / AMR
    • Movable Rack
      • Robotic mobile fulfillment (RMF) system
    • Static Rack
      • Pick support AGVs
      • (Mobile) Pick AGVs
  • Pick stations
    • Manual picking
    • Robot picking

AMR vs. AGV | What you should know | Storage Solutions (storage-solutions.com)

While on the surface the two may seem very similar, there are several key factors that set an AGV and an AMR apart. The most fundamental difference between the two comes down to how they move through a facility. A guided vehicle is just what the name implies, a vehicle that operates on a fixed and guided route. Usually AGVs run along some sort of path, such as embedded magnets or wires. With this set path, an AGV does not have the ability to make quick changes on the fly, and does not always have the intelligence to stop when something comes in its path. This can lead to expensive upgrades down the road as your operational needs change. It could also put you at risk for more work-related accidents to occur.

An AMR is an autonomously navigating robot designed to work collaboratively with humans. Therefore it offers a much smarter solution. With an AMR, you get a flexible solution that offers easy adjustments as your operational needs change. In addition, the implementation for an AMR is much simpler. There is no need to make costly facility adjustments to get started. Finally, another key point that sets an AMR apart is the safety factors. AMRs use vision systems and laser sensors to ensure that accidents are avoided in any environment.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 918–19).

Automated picking systems

Classification by Azadeh et al. (2019)

Shuttle (AVS/R)

• A shuttle is sometimes referred to as an automated guided vehicle (AGV) or robot.
• Shuttle-based storage and retrieval system = autonomous vehicle-based storage and retrieval (AVS/R).

• Types:
  • Aisle-based shuttle systems.
  • Grid-based shuttle systems.

“During the last decade, warehouse automation has developed rapidly. A big boost to this development has been given by autonomous vehicle–based storage and retrieval (AVS/R), or shuttle-based storage and retrieval, systems. These systems use racks with aisles and deploy autonomous shuttles that operate at each level in each aisle. Vertical transport is enabled by lifts.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 917).

Automated picking systems: Shuttle (AVS/R)

Aisle-based shuttle systems

• Throughput capacity of AS/R systems is constrained:
  • One crane handles loads at all vertical levels within a given storage aisle.

• Autonomous vehicle–based storage and retrieval (AVS/R) systems:
  • require investment similar to AS/R systems,
  • offer a much higher retrieval capacity than AS/RS, and
  • are more flexible in capacity (easy to add or remove shuttles).

• AVS/RS classification based on shuttles’ movement capability:
  • horizontal,
  • vertical, and
  • diagonal.

“Aisle-Based Shuttle Systems

Throughput capacity of AS/R systems is constrained because only one crane is responsible for handling loads at all vertical levels within a given storage aisle. This led to a new generation of automated orderpicking systems, autonomous vehicle–based storage and retrieval systems, which were first introduced by Savoye Logistics in the 1990s. Such systems are increasingly popular because the required investment is similar to that of AS/R systems, although they offer a much higher retrieval capacity and are also significantly more flexible in capacity. By using additional shuttles, system capacity can be increased, and by removing shuttles, capacity can be decreased. Typical AVS/R systems use shuttles, which can drive in the x direction and the y direction on any level in the aisle, and lifts move shuttles (or unit loads) between the levels. In this variant, shuttles can move only horizontally, and rely on lifts for vertical movements. Recently, several robotic solutions have emerged, in which the shuttles (called robots) have the ability to not only move horizontally but also elevate up to different tiers by either moving diagonally or vertically (Azadeh, Roy, and De Koster 2018). Therefore, the AVS/R system can be classified based on their shuttles’ movement capability into three categories: horizontal, vertical, and diagonal systems (see Figure 2). In this section, we discuss different types of horizontal systems and leave the discussion on vertical and diagonal systems for Section 8.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 926).

Automated picking: Aisle-based shuttle systems

Single/Double-deep storage

• Shuttles can reach loads in one- and two-deep lanes.

The storage area in an AVS/R system consists of aisles with multitier storage racks on both sides and a cross aisle that runs orthogonal to the aisles. To perform storage and retrieval actions, a lift is used for vertical movements between tiers and autonomous vehicles or shuttles are used for the horizontal movements within the tier. To retrieve a tote, a shuttle moves to the tote’s storage location and picks up the tote, pulls it on board and moves toward the lift for vertical travel. Then the shuttle either hands the tote to the lift (tier-captive system), or uses the lift to move the load to a lower level (tier-to-tier system) where it is transferred to the pick station by conveyor belt. After picking, the tote again uses the lift and a shuttle to be stored in the system.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 5.1.1, p. 926).

Automated picking: Aisle-based shuttle systems (single/double deep)

Horizontal

Figure 3: Shuttle-based AVS/R system with built-in sorting and sequencing capabilities. A lift is used for vertical movements between tiers and autonomous vehicles or shuttles are used for the horizontal movements within the tier (ADAPTO - Vanderlande)

The storage area in an AVS/R system consists of aisles with multitier storage racks on both sides and a cross aisle that runs orthogonal to the aisles (see Figure 9). To perform storage and retrieval actions, a lift is used for vertical movements between tiers and autonomous vehicles or shuttles are used for the horizontal movements within the tier (Roy 2011). To retrieve a tote, a shuttle moves to the tote’s storage location and picks up the tote, pulls it on board and moves toward the lift for vertical travel. Then the shuttle either hands the tote to the lift (tier-captive system; Heragu et al. 2008), or uses the lift to move the load to a lower level (tier-to-tier system; Heragu et al. 2008) where it is transferred to the pick station by conveyor belt. After picking, the tote again uses the lift and a shuttle to be stored in the system.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 5.1.1, p. 926).

ADAPTO is our intelligent shuttle-based automated storage and retrieval system (AS/RS) with built-in sorting and sequencing capabilities.

— Vanderlande Systems (Vanderlande, n.d., ADAPTO product card).

The named sorting and sequencing capabilities are specific to Vanderlande’s ADAPTO product, while the journal excerpt explains the general horizontal shuttle mechanism.

Automated picking: Aisle-based shuttle systems (single/double deep)

Vertical

Figure 4: The Perfect Pick system uses robots (iBOTs) to perform storage and retrieval operations. The system can be extended by adding modules, aisles, or robots (OPEX Perfect Pick - Bastian Solutions)

Figure 5: Robots climb up to 12 m and move in three dimensions to transport bins/trays between racks and picking stations (Skypod - Exotec).

In these systems, a single robot can independently roam the storage rack to perform storage and retrieval operations (no lift is required). In a diagonal system, robots also move diagonally, and in a vertical system, robots also move vertically inside the rack structure to elevate to the upper levels. The RackRacer developed by Fraunhofer Institute for Material Flow and Logistics is an example of a diagonal system. Perfect Pick developed by OPEX Corporation and the Skypod developed by Exotec Solutions are two examples of vertical systems. The Perfect Pick system uses robots, called iBots, to perform storage and retrieval actions.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, p. 936).

The current OPEX product page states that Perfect Pick uses wireless iBOTs to find, retrieve, and deliver totes, with a maximum system height of 11.7 m (OPEX Corporation, n.d., product description and technical specifications). The Exotec/CEVA release separately states that Skypod floor robots climb through racks up to 12 m (CEVA Logistics and Exotec 2023, 1). These are two distinct vendor specifications, not one general height limit for vertical systems.

Automated picking: Aisle-based shuttle systems (single/double deep)

Diagonal

Figure 6: A climbing shuttle moves autonomously in horizontal and diagonal directions — without a lift or additional rails — using “crawler elements”, which climb supporting spots (RackRacer - Fraunhofer Institute for Material Flow and Logistics)

In these systems, a single robot can independently roam the storage rack to perform storage and retrieval operations (no lift is required). In a diagonal system, robots also move diagonally, and in a vertical system, robots also move vertically inside the rack structure to elevate to the upper levels. The RackRacer (see Figure 17(a)) developed by Fraunhofer Institute for Material Flow and Logistics (2014) is an example of a diagonal system. Perfect Pick® developed by OPEX Corporation (2013) and the SkypodTM (see Figure 17(c)) developed by Exotec Solutions (2017) are two examples of vertical systems. The Perfect Pick system uses robots, called iBots® (see Figure 17(b)), to perform storage and retrieval actions (Azadeh, Roy, and De Koster 2018).

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, p. 936).

An automated shuttle warehouse without a lift or other expensive customization seemed to be impossible - until now. The Fraunhofer Institute for Material Flow and Logistics presents an innovative solution to this problem: the »RackRacer«. The new climbing shuttle moves autonomously in horizontal and diagonal directions - without a lift or additional rails. This process is enabled by so-called »crawler elements« which support themselves on specific climbing or supporting spots.

— RackRacer (Fraunhofer Institute for Material Flow and Logistics, n.d., heading “The idea”).

Fraunhofer explains that crawler elements brace against specific supporting points, enabling the named shuttle to reach rack positions without indirect routes (Fraunhofer Institute for Material Flow and Logistics, n.d., heading “The idea”). This evidence is specific to RackRacer, not every diagonal system.

Automated picking: Aisle-based shuttle systems

Multi-deep storage

• Multi-deep storage lanes are widely used in the food and beverage industry.
  • Large volumes of products of limited variety.
  • Each lane contains unit loads of the same item (possibly a single batch or production lot).

• In a multi-deep AVS/R system:
  • Each tier has an AV¹ that consists of:
    • Shuttle: Moves along the main cross aisle.
    • Satellite: Transports the unit load.

Figure 7: Unit-load autonomous vehicle storage and retrieval system (AVS/R) with multiple deep lane storage (Manzini et al. 2016)

“We introduce the term module as a group of lanes on one side of the aisle. Each tier has an autonomous vehicle that consists of a shuttle and a satellite: the shuttle moves back and forth along the main cross aisle, transporting the satellite. The satellite transports the palletised unit load. The satellite is the main storage and retrieval (S/R) device. The satellite departs from the shuttle and enters the storage lanes to store and retrieve unit loads. Inside a lane, the pallets are stored along the channel. The satellite stops exactly in front of the last previously inserted palletised unit load. In this location, the satellite performs the next retrieving or stocking action applied to this SKU and batch, which is in agreement with the last-in-first-out (LIFO) policy for put-away activities. Figure 1 summaries the adopted notation.”

— Travel Time Models for Deep-Lane Unit-Load Autonomous Vehicle Storage and Retrieval System (AVS/RS) (Manzini et al. 2016, article p. 1, PDF p. 3).

1. AV: Autonomous Vehicle

Automated picking: Aisle-based shuttle systems (multi-deep)

Multi-deep shuttle/transfer car

Figure 8: FIFO can be achieved by providing a shuttle lane on both sides of the storage rack (Multi Deep Shuttle ASRS with Robotic Palletizer/Depalletizer - DAIFUKU).

Finished Product Warehouse for Beverage Factory

In the soft drink manufacturing industry, the top SKUs often account for about 80% of the total volume, and the amount of inventory per SKU is higher than in other industries. The Shuttle D3 is the ideal fit for this kind of small-mix, high-volume storage need.

In the above model system, beverage product cases transported from the production lines are stacked on pallets by a layer palletizer (1), and then entered into the D3 (4) via a pallet conveyor (2) and a high-speed rail-guided vehicle STV (3). Pallets that have been ordered for shipment are sent out from the D3 via an STV to the truck berth side conveyor (5). Stacked, empty pallets returned to the site are fed from the truck berth side conveyor (6) through the STV to the layer palletizer. Combined with other material handling equipment, the D3 provides high-level automation from production to shipping.

— Multi-Deep Shuttle Storage System “D3” (“Multi-Deep Shuttle Storage System ‘D3’ | Automated Storage ＆ Retrieval System (AS/RS) | Products | Intralogistics | Solutions,” n.d., sec. “Finished Product Warehouse for Beverage Factory”).

Automated picking

Grid-based shuttle systems

• Automated storage and retrieval system where shuttles move on a grid.
• Storage locations are either
  • Dynamic (or puzzle-based system): The stored SKUs move on shuttles to store or retrieve items.
  • Static: AVs move the stored SKU.

“Grid-Based Shuttle Systems Grid-Based Shuttle Systems In this section, we discuss a variant of the shuttle-based automated storage and retrieval systems in which shuttles move on a grid. In a grid-based system, the storage locations are either dynamic or static (see Figure 2). In a dynamic storage system (or puzzle-based system), the stored SKUs need to move (on a shuttle) to store or retrieve an item. We discuss the static storage systems in Section 8.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 929).

Automated picking: Grid-based shuttle systems

Dynamic storage (puzzle-based)

• Concept based on Sam Loyd’s 15-tile puzzle.
• A tile represents a load (e.g., a tote, pallet, or container) stored in a grid with one open location.
• Storage density comparison:
  • \(k\)-deep aisle-based system: less than or equal to \(2k/(2k + 1)\). Example:
    • single-deep = \(2/3\).
    • double-deep = \(4/5\).
  • \(n\)-cell grid: \((n − 1)/n\).

• To retrieve a requested unit load, the system repeatedly moves the open locations (called “escorts”) to bring the load to the I/O point.

Figure 9: Open location “escorts” the requested item to the I/O point (Azadeh et al. 2019)

“System Description Gue (2006) shows that the storage density of a k-deep aisle-based system is less than or equal to 2k/(2k + 1), that is, 2/3 for a single-deep and 4/5 for a double-deep system. To achieve an absolute maximum storage density, a new concept based on Sam Loyd’s famous puzzle game has been developed; the 15-slide puzzle (Loyd 1914). The 15-slide puzzle is a game in which 15 numbered tiles slide within a 4 × 4 grid, and the objective of the game is to arrange the tiles in the correct numerical sequence, starting from a random initial arrangement. The puzzle-based storage and retrieval concept (Gue and Kim 2007), follows a similar idea. A tile represents a tote, a pallet, or even a container that is stored in a grid with only one open spot on the grid, which allows a (n − 1)/n storage density, where n is the number of cells in the grid. To retrieve a requested unit load, the system repeatedly moves the open locations, which ultimately brings the load to the I/O point. This is illustrated in Figure 11. To retrieve a load, an open location first needs to be moved next to the requested item. Then the open location should be used to move the item to the I/O point. In other words, the open location”escorts” the requested item to the I/O point. An open location is called an escort (Gue and Kim 2007). Several compact storage system variants have emerged from the puzzle-based concept in practice and in the literature.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 929, PDF p. 17).

Automated picking: Grid-based shuttle systems (Dynamic storage)

GridStore

• High-density storage system for physical goods
  • Rectangular grid of square conveyor modules:
    • Modules move items in the four cardinal directions.
    • Modules communicate with their neighbors and with the item they carry.

• Throughput can be controlled by:
  • Number of rows and columns in the grid.
  • Number of empty cells \(k\).
  • Number of requested items = work-in-process (WIP).

Figure 10: At the north side, a replenishment conveyor moves replenishment items (blue) in the grid. At the south side, the retrieval conveyor moves requested items (black) away from the grid (GridStore)

a rectangular grid of square conveyor modules with the capability to move items in the four cardinal directions. The modules can communicate with their neighboring modules as well as with the item they carry. At the south side of the grid, the retrieval conveyor moves products away from the grid. At the north, a replenishment conveyor moves products that need to be stored in the grid.

“GridStore Building on the puzzle-based storage system concept, Gue et al. (2014) propose a high-density storage system for physical goods called GridStore. The system consists of a rectangular grid of square conveyor modules with the capability to move items in the four cardinal directions. The modules can communicate with their neighboring modules as well as with the item they carry. At the south side of the grid, the retrieval conveyor moves products away from the grid. At the north, a replenishment conveyor moves products that need to be stored in the grid. Figure 12 illustrates the movement of the items toward the retrieval conveyor.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 929).

GridStore—a Grid-Based Storage and Retrieval System

After many years of development and testing, we finally have a completed and accepted paper on GridStore: A Puzzle-Based Storage System with Decentralized Control. The paper, which will appear in IEEE Transactions on Automation Science and Engineering, was written with my colleague and friend Kai Furmans from Karlsruhe Institute of Technology and with two of our students, Zäzilia Seibold and Onur Uludağ.

For those familiar with this work from presentations gone by, you will notice that the name of the system has changed from GridFlow to GridStore. We are beginning to use the term GridFlow to refer to the class of systems that uses a grid-based movement architecture, rather than to a specific instance. GridFlow, GridPick, and GridSequence are all instances of the GridFlow way of thinking.

Below is an example of GridStore at work. Gray cells contain an item, such as a tote or pallet. Black cells represent requested totes trying to make their way to the southern boundary, where they are removed from the system by a conveyor or other means. Gray cells with a black circle represent totes returning to their “home rows.” When a replenishing totes reaches its home row, it turns gray until requested, after which it turns black. White cells are empty.

At each iteration conveyor cells determine their current state and then engage their neighbors in two rounds of “negotiations,” in which cells with competing interests try to reconcile. After negotiations are complete, cells convey in the appropriate direction. Voilà—GridStore!

The main feature of this paper is a detailed description of the decentralized controls rules that effect the nice global behavior of GridStore. Describing these rules was no easy task, so feel free to contact us if the paper doesn’t make sense.

GridStore

Our early research on high-density storage system suggested a natural tradeoff between storage density and throughput—the higher the density, the lower the throughput. Is there a way to get both?

The GridStore system combines puzzle-based movement and decentralized control to produce a high-density storage system that conforms to the plug-and-play paradigm. The system is built on the same hardware model as the Flexconveyor—a unit-sized conveyor segment that conveys in the four cardinal directions, has communications ports on all four sides, and has its own controller. Each conveyor is capable of talking to its four neighbors, and no others.

Control of the GridStore system is completely decentralized. At each time step, conveyors assess their current state (Unoccupied? Occupied? Occupied with moving item?), then enter a round of electronic negotiations with their neighbors to determine which direction each conveyor will convey. The result is the global behavior you see here:

Black items in the video have been requested by an external source, such as an order picking workstation or a shipping schedule. Blue items are returning to replenish the system. Each blue item has a target row, to which it travels and then awaits a request. Target rows match the home rows of departing items, thereby maintaining system balance.

Throughput in the GridStore system can be controlled by the number of rows and columns in the grid, by the number of empty cells k, and by the number of requested items (WIP), which we keep constant in a sort of “constant work-in-process” (CONWIP) system. Throughput goes up as the number of empty cells increases, and it also goes up as WIP increases. In ordinary storage systems subject to congestion, the marginal increase in throughput goes down as WIP increases; the opposite is true in GridStore, for low to medium levels of WIP.

As the number of active requests (WIP) increase, throughput increases non-linearly because (1) additional WIP tends to empty out the lower rows, clearing them for requested items and (2) requested items are more likely to form “slugs” which move together very efficiently. For very high levels of WIP (60+ in the plot), most items are requested and the system becomes “river-like.”

If you would like to learn more about the algorithms behind GridStore, please email me with the form below.

Automated picking: Grid-based shuttle systems (Dynamic storage)

GridPick

• High-density storage container system:
  • No fixed lanes/aisles.
  • Open spots on the grid allow items to move during the retrieval process.
  • Only containers holding requested items move to the edge of the system (pick face).
• Unlike the GridStore, items do not leave the grid.

Figure 11: The picker picks the items and accumulates the order in a picking cart. The next order is released when all items from the current order have arrived at the system’s pick face.

A backward movement balance the empty cells in each row.

“GridPick Based on the GridStore architecture, Uludag (2014) introduces an order-picking systems called GridPick. The system is filled with high-density storage containers, without any fixed lanes or aisles; only a few open spots on the grid allow items to move during the retrieval process. The objective of the system is to provide a high order-picking rate while minimizing any congestion effects. Unlike the GridStore, items do not leave the grid in the GridPick system. Only containers holding the requested item move to the edge of the system, called the pick face. The picker picks the items and accumulates the order in a picking cart. There is also a backward movement, away from the pick face, to balance the empty cells in each row. This balancing rule helps to avoid deadlocks in the system. Figure 13 illustrates an instance of GridPick. The gray items are not-requested stored items, and the black items are the requested items that are moving toward the pick face. Black circles on top of some gray items are balancing items moving in the opposite direction from the pick station. The numbers on top of the items display the order number for the requested item. The next order for picking is released when all the items from an order have arrived in the pick face of the system. When comparing the GridPick with its equivalent gravity flow rack counterpart, Gue and Uludag (2012) show that the gravity flow rack results in a larger system. Therefore, the average productivity (measured in picks per hour) is higher for smaller orders in the GridPick system because it reduces travel. However, as order size increases, walking time of both systems converges to the same number.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 929).

Automated picking: Grid-based shuttle systems (Dynamic storage)

Live-Cube

• Multilevel system in which each floor is based on a puzzle-based storage architecture.
• Main components:
  • Multiple levels of storage grids.
  • Shuttles (can move autonomously and simultaneously).
  • Lifts.
  • I/O points.
  • Shuttles move in \(x\) and \(y\) directions with the load on top of them.
  • Loads (individually accessible).

Figure 12: Live-cube-based system with I/O point at the ground level at the lower left corner (Zaerpour et al. 2017)

“Live-Cube Compact Storage A multilevel system in which each floor is based on a puzzle-based storage architecture is called a live-cube storage system (Zaerpour, Yu, and De Koster 2015). As illustrated in Figure 14, the essential parts of the system are multiple levels of storage grids, shuttles, lifts, and the I/O points. Each level of the system forms a grid-based storage system where shuttles move in x and y directions with the load on top of them. With at least one escort available in each level, the shuttles maneuver the requested item to the lift, which transports the load to the I/O point. The I/O point is usually located at the lower left corner of the system.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 930).

— Small Is Beautiful: A Framework for Evaluating and Optimizing Live-Cube Compact Storage Systems (Zaerpour et al. 2017, figure on article p. 2, PDF p. 3).

Automated picking: Grid-based shuttle systems (Dynamic storage)

GridFlow

• Instead of conveyors (puzzle-based system), AGVs are used to move the pallets.
• Increased flexibility:
  • In a conventional puzzle-based system, the physical layout cannot be changed easily.
  • In the GridFlow system, vehicles can form grids of any shape.

“GridFlow System A major drawback of the puzzle-based system is that the physical layout cannot be changed easily. Therefore, the concept of GridFlow is proposed by Furmans, Nobbe, and Schwab (2011) to offer a cheaper and a more flexible system. In this system, instead of conveyors, AGVs are used to move the pallets. The use of AGVs instead of conveyors makes the system more flexible with respect to design and throughput changes. Vehicles can form grids of any shape without any additional investment. Figure 15 illustrates the GridFlow system and the vehicle movements in the system. Many system manufacturers are developing puzzlebased systems in different variants. The number of actual implementations and prototypes based on this concept is growing in many different fields, especially in automated parking systems (e.g.,”park, swipe, leave” parking system, Automation Parking Systems 2018; Hyundai integrated parking system, Hyundai Elevator Co. 2016;W ̈ ohr Parksafe, W ̈ ohr 2016).”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 930).

Automated picking: Grid-based shuttle systems (Dynamic storage)

GridSort

• Based on the GridFlow system.
• Transport and sort loads using either:
  • Modular unit-sized four-directional conveyors (FlexConveyors).
  • AGVs.

Figure 13: An AGV fleet on a grid sorts parcels by destination (Sorting Robot - Libiao Robotics)

“GridSort The GridSort system is based on the GridFlow system discussed in Section 6. It uses modular four-directional conveyors, called FlexConveyors (Furmans, Sch ̈ onung, and Gue 2010), or AGVs to transport and sort the loads. Recently, Libiao Robotics developedadifferenttypeofGridSortsystem,used by several parcel carriers in China, that uses a fleet of hundreds of autonomous AGVs on a grid to sort parcels by destination.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 936).

“GridSort differs fundamentally from conventional conveyor-based sorters and new models will be required to evaluate its performance. For example, in GridSort the movements of the shuttles (that carry the loads) depend on the empty spaces on the grid. So a research question may be, how can the empty spaces on the grid be exploited to simultaneously move multiple loads in the system efficiently?”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 938).

Automated picking: Grid-based shuttle systems

Static storage

• Grid-based system in which items are stored in a very dense storage stack with a grid on top.
• In each cell of the grid, bins that contain the items are stacked on top of each other and form the storage stacks.

The robot-based compact storage and retrieval (RCSR) system is another type of grid-based system (see Section 6) in which items are stored in a very dense storage stack with a grid on top (see Figure 18). In each cell of the grid, bins that contain the items are stacked on top of each other and form the storage stacks. The workstations are located at the lowest level next to the storage stacks. Robots roam on top of the storage block on the grid. The robots have lifting capabilities and can extract bins from the storage frames and transport them to the workstations (Zou, De Koster, and Xu 2016). AutoStoreTM, developed by Hatteland is the first implementation of an RCSR system (see AutoStore 2018). Recently the British retailer Ocado developed a similar system. Zou, De Koster, and Xu (2016) are the only ones to investigate the RCSR system. They model the system as a semiopen queueing network and compare two storage policies, namely, dedicated and shared storage. They show that the dedicated policy results in a shorter throughput time, whereas the shared policy has financial benefits because of substantial cost savings in the total storage space. They also optimize the shape of the system and show that the width-to-length ratio is around 2/3 when using random storage stacks, and slightly larger when using zoned storage racks. They also show that immediate reshuffling can improve the dual command throughput time compared with delayed reshuffling.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, pp. 936–937).

The layout and configuration of each AutoStore are customized to fit your warehouse and specific fulfillment needs. It achieves high-speed efficiency by being designed to execute and optimize many smaller tasks all at the same time, for example fetching multiple Bins at simulataneously to retrieve different line items for different orders. The more modules an AutoStore is configured with (more Ports & Robots), the higher the maximum potential throughput it can reach, as it achieve a higher overall aggregated output.

Items are stored inside containers called Bins, stacked tightly in rows inside the Grid. Retrieval is automated via a fleet of Robots on top of the Grid retrieving Bins and transporting them to workstations referred to as Ports.

— AutoStore Solution FAQs (AutoStore, n.d., heading “How it works”).

There are 76 Robots at the main warehouse, which also accommodates 22,933 Bins, and the grid covers roughly 20,000 square feet (1,860 square meters). There are 11 Ports for storage and retrieval.

“As Hayabusa’s required storage and retrieval capacity is very high, we expanded the area of the AutoStore grid by making the picking platform four Bins high in certain areas and 14 Bins high in others,” Shida notes. “This sorts out the problem of the retrieval time for products at the very bottom.”

A customized layout makes it possible for four levels of Bins to be stored above the employees’ heads. These four levels contain 60% of the most frequently ordered products so that the Robots don’t have to dig deeper than necessary to retrieve the products.

— Warehouse Automation Boosts Hayabusa (AutoStore 2025, heading “How AutoStore works at Hayabusa”).

The shallow fast-product arrangement is evidence from the Hayabusa case, not a generic performance guarantee for every RCSR installation.

Automated picking: Grid-based shuttle systems (Static storage)

Robotic compact S/R system

• Robot-based compact storage and retrieval (RCSR) system:
  • Robots roam on top of the storage block on the grid and
  1. extract bins from the storage frames (lift);
  2. transport them to workstations.
  • The workstations are located at the lowest level next to the storage stacks.

Figure 14: “AutoStore” promises 40%–60% space reduction and allows capacity increase by adding robots, workstations, and grid (bin locations) (AutoStore - Bastian Solutions)

The robot-based compact storage and retrieval (RCSR) system is another type of grid-based system in which items are stored in a very dense storage stack with a grid on top. In each cell of the grid, bins that contain the items are stacked on top of each other and form the storage stacks. The workstations are located at the lowest level next to the storage stacks. Robots roam on top of the storage block on the grid. The robots have lifting capabilities and can extract bins from the storage frames and transport them to the workstations. AutoStore, developed by Hatteland, is the first implementation of an RCSR system. Recently the British retailer Ocado developed a similar system. Zou, De Koster, and Xu model the system as a semi-open queueing network and compare two storage policies, namely, dedicated and shared storage. They show that the dedicated policy results in a shorter throughput time, whereas the shared policy has financial benefits because of substantial cost savings in the total storage space. They also optimize the shape of the system and show that the width-to-length ratio is around 2/3 when using random storage stacks, and slightly larger when using zoned storage racks. They also show that immediate reshuffling can improve the dual command throughput time compared with delayed reshuffling.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, pp. 936–937).

Toyota Material Handling Nederland states that this AutoStore configuration can save 40%–60% of space and can be scaled by adding robots, workstations, and grid capacity (Toyota Material Handling Nederland B.V., n.d., heading “Eigenschappen en voordelen AutoStore systeem”). This is a current manufacturer-group claim, not a general research result for every RCSR installation.

The embedded video URL is identified in the U.S. Patent and Trademark Office record as AS-1013, AutoStore Logistic-Technical Presentation, dated 2009-08-26 (U.S. Patent and Trademark Office, Patent Trial and Appeal Board 2022, exhibit AS-1013).

Automated picking

Automated Guided Vehicle (AGV)

• AGVs or autonomous mobile robots (AMRs) support or carry out the picking process.

• System configurations:

  • Movable rack
  • Static rack

The slide uses AGV/AMR as an umbrella for two distinct system families.

Internet retailers typically have warehouses with large assortments of small products. Their demands usually consists of multiline small-quantity orders. In manual picking systems, much nonvalue added time is needed by the pickers to travel along the aisles. The robotic mobile fulfillment (RMF) system is a system in which robots capable of lifting and carrying movable shelves retrieve the storage pods (i.e., movable shelf racks) and transport them to the pickers, who work in ergonomically designed workstations. Bringing the inventory to the picker instead of the picker traveling to the inventory can double the picker productivity (Wurman, D’Andrea, and Mountz 2008). The system is also very flexible in throughput capacity, as more robots and pods can be added to the warehouse. This is particularly important for internet retailers who face volatile demand.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 7, pp. 932–933).

Most retail warehouses still use manual order-picking systems. Retail stores usually place large replenishment orders at the distribution center. The distribution center then ships the orders in multiple roll cages or on pallets. Therefore, a single order requires multiple pick tours (trips between pick locations and the depot). Recently, AGV-based pick systems, called pick support AGV systems, have been developed to minimize the picker travel time to fill large orders (see Figure 19). In these systems, an AGV automatically follows the picker closely and transports the roll cages, so that the picker can drop off the retrieved items. Once the roll cage is full, the AGV is automatically swapped with a new AGV carrying an empty roll cage. The picker can continue the picking route without returning to the depot, and the AGV automatically transports the full roll cage to the depot.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, pp. 936–937).

RMF moves inventory pods to stationary pickers; pick-support AGVs remain with static racks and move containers alongside or between human pickers. The two families should not be treated as the same mechanism.

Automated picking: Automated Guided Vehicle

Movable Rack

• Internet retailers typically face demand for multi-line, small-quantity orders.
  • Travel along the aisles = non-value-added time.
• Bringing the inventory to the picker instead of the picker traveling to the inventory can double picking productivity.

Internet retailers typically have warehouses with large assortments of small products. Their demands usually consists of multiline small-quantity orders. In manual picking systems, much nonvalue added time is needed by the pickers to travel along the aisles. The robotic mobile fulfillment (RMF) system is a system in which robots capable of lifting and carrying movable shelves retrieve the storage pods (i.e., movable shelf racks) and transport them to the pickers, who work in ergonomically designed workstations. Bringing the inventory to the picker instead of the picker traveling to the inventory can double the picker productivity (Wurman, D’Andrea, and Mountz 2008). The system is also very flexible in throughput capacity, as more robots and pods can be added to the warehouse. This is particularly important for internet retailers who face volatile demand.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 7, pp. 932–933).

The review attributes the doubling result to Wurman, D’Andrea, and Mountz (2008); it is a reported result, not a universal productivity guarantee.

Automated picking: Automated Guided Vehicle (Movable rack)

Robotic mobile fulfillment (RMF) system

• Major components:
  • Robotic drive units: robots transporting inventory pods to the workstation for restocking/picking.
  • Inventory pods: movable shelf racks with stored products.
  • Workstations: ergonomically designed areas where human workers perform pod replenishment, picking, and packing functions.
• Flexible in throughput capacity:
  • More robots and pods can be added to the warehouse.
  • Important for internet retailers who face volatile demand.

Figure 15: Amazon RMF system features AGVs (Kiva) and, more recently, AMRs (Proteus) (Amazon robotics)

“7.1. System Description The RMF system consists of three major components: 1. Robotic drive units. These robots are instructed by the central computer to transport inventory pods to the workstation for restocking or for picking. Nowadays also decentrally (or locally) controlled systems exist. 2. Inventory pods. Pods are movable shelf racks that contain the stored products. Pods come in two standard sizes. Smaller pods are used for weights up to 450 kg, and large pods are used for weights up to 1300 kg. Workstations. Workstations are ergonomically designed areas where human workers perform pod replenishment, picking and packing functions (MWPVL International 2012). Figure 16 presents an RMF system workstation and its components. To pick an ordered item with the RMF system, the order is first assigned to one of the workstations. Then the item is assigned to a pod and one robot. The robot then moves from its dwell location to retrieve the pod. At this point, the robot moves without a load and can therefore move underneath the pods, without using the designated travel aisles. Once the robot reaches the desired pod, it moves underneath it, lifts the pod, and transports it to the workstation via the travel aisles. The robot enters the workstation buffer and waits for its turn (see Figure 16). The picker takes the requested products and adds them to the customer order bin placed in a different rack. The robot then returns the item pod to a storage location that accounts for the frequency of the requests for the pod. The storage locations are therefore fully dynamic (Wurman, D’Andrea, and Mountz 2008; Enright and Wurman 2011). The layout can be fully adapted both dynamically and automatically to the product and order characteristics.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 932).

Automated picking: Automated Guided Vehicle

Static Rack

• AGVs are used to pick items at racks:
  • Autonomously;
  • In parallel with a picker (human-machine interaction).
• Popular in practice because of simplicity and flexibility.

“The differentiating characteristic of pick support AGVs is the collaboration between human pickers and AGVs. The parallel movement of pickers and AGVs makes the modeling, analysis, and optimization of this system completely different from manual picking systems or the robotic systems mentioned earlier in this paper. Evaluating the performance of these systems is an interesting stream for future research. The interesting research question is, how can we coordinate the parallel movement of the pickers and AGVs to maximize the throughput?”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 938).

“Human picking in collaboration with AGVs is one of the most recent technologies that is becoming popular in practice because of its simplicity and flexibility, but it has not yet been adequately studied.”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 940).

Automated picking: Automated Guided Vehicle (Static rack)

Pick support AGVs/AMRs

• Challenges in retail warehouses:
  • Single order requires multiple pick tours (trips between pick locations and the depot).
  • Large replenishment orders.
  • Orders shipped in multiple roll cages/pallets.
• Pick-support AGV systems minimize the picker travel time to fill large orders:
  • AGV automatically follows the picker and transports the roll cages.
  • Picker can continue the picking route without returning to the depot.
  • Once the roll cage is full:
    • The AGV is automatically swapped with a new AGV carrying an empty roll cage.
    • AGV automatically transports the full roll cage to the depot.

Most retail warehouses still use manual order-picking systems. Retail stores usually place large replenishment orders at the distribution center. The distribution center then ships the orders in multiple roll cages or on pallets. Therefore, a single order requires multiple pick tours (trips between pick locations and the depot). Recently, AGV-based pick systems, called pick support AGV systems, have been developed to minimize the picker travel time to fill large orders (see Figure 19). In these systems, an AGV automatically follows the picker closely and transports the roll cages, so that the picker can drop off the retrieved items. Once the roll cage is full, the AGV is automatically swapped with a new AGV carrying an empty roll cage. The picker can continue the picking route without returning to the depot, and the AGV automatically transports the full roll cage to the depot.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, pp. 936–937).

The travel reduction comes from separating the picker’s aisle work from the container’s return trip to the depot.

Automated picking: Automated Guided Vehicle (Static rack)

Pick support AGV example

Figure 16: “Chuck” is an AMR that can be used in all put-away, picking, counting, replenishment and sorting tasks Collaborative Mobile Robot - 6 River Systems

Most retail warehouses still use manual order-picking systems. Retail stores usually place large replenishment orders at the distribution center. The distribution center then ships the orders in multiple roll cages or on pallets. Therefore, a single order requires multiple pick tours (trips between pick locations and the depot). Recently, AGV-based pick systems, called pick support AGV systems, have been developed to minimize the picker travel time to fill large orders. In these systems, an AGV automatically follows the picker closely and transports the roll cages, so that the picker can drop off the retrieved items. Once the roll cage is full, the AGV is automatically swapped with a new AGV carrying an empty roll cage. The picker can continue the picking route without returning to the depot, and the AGV automatically transports the full roll cage to the depot.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, sec. 8.3.1, pp. 936–937).

Azadeh et al. support the system class. The 6 River Systems brochure identifies Chuck within workflows for picking, packing, sorting, and replenishment (6 River Systems 2020, fulfillment-workflow section). The current product page also lists put-away and counting as applications (6 River Systems, n.d., applications heading). These latter claims are vendor-specific.

Automated picking: Automated Guided Vehicle (Static rack)

(Mobile) Pick AGVs

Figure 17: An AGV that automatically goes to the picking location and picks the item. Once the order is complete, picked items are transported to the depot (Zalando’s smart intralogistics - Magazino)

“Pick Support AGVs Locus Robotics (2018) has developed another variant of this system. Instead of following the picker, their AGV (called LocusBotsTM) automatically goes to the pick location and waits for the picker to arrive. Once the picker puts the item into a customer tote carried by the AGV, the AGV goes to the next location. When the order is complete, the AGV transports the tote to the depot. Some systems automate the whole picking process. An example is the TORUTM picking robot. In this variant, the AGV automatically goes to the picking location and picks up the item without any help from the picker. Similar to the previous variants, once the order is complete, the AGV transports the picked items to the depot (see Magazino 2017).”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 936).

Automated picking

Pick stations (manual vs. robot picking)

• Most automated picking systems are combined with manual piece picking at picking stations.
• Efficient robot picking depends on technologies such as:
  • deep learning,
  • rapid image processing, and
  • appropriate gripper selection.

Human-machine interaction: humans may still be needed to collaborate with robots at picking stations in the coming decades.

Figure 18: Using deep learning to pick up groceries with a robotic arm (Item Picking - Fizyr).

“Analyzing Robot Pick Stations Most automated systems described in Figure 2 and the previous sections retrieve unit loads and bring them to an order pick station. Even with new technologies, like RCSR systems, puzzle-based systems, or RMF systems, piece picking at the station is done manually (De Koster and Yu 2008; Füßler and Boysen 2017, 2019). However, new technologies, like deep learning and rapid image processing, are developing that make automated product recognition, selection of the appropriate gripper, and rapid automated picking with robots possible. For the coming decade or so, robot picking stations alone cannot do the job cost-efficiently and at sufficient speed, which means humans may have to collaborate with robots. This calls for further study of human–machine interaction”

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, 938).

Depalletizing system

Figure 19: Bag depalletizing (Fizyr)

Fizyr’s current vendor case describes AVT’s heavy-bag depalletizing solution as a robotic pick-and-place application using Fizyr Vision AI. This evidence identifies the example and its vision component; it is not comparative research evidence about all depalletizing systems (Fizyr, n.d., AVT Heavy Bag Depalletizing subsection).

Autonomous inventory drones

Figure 20: Inventory scanned by drones can be compared with information in the WMS (DSV).

DSV’s primary release describes autonomous drones that scan warehouse pallet barcodes and occupancy, operate outside normal working hours, and compare the scan results with information in DSV’s systems. The slide is therefore a case of DSV’s implementation, not a general performance claim for every inventory-drone system (DSV 2020, main release).

Takeaways

• Why automation?
  • Potential savings on space, labor, and operating costs.
  • 24/7 availability.
  • Scalability and throughput flexibility.

Caveat: ROI is medium- to long-term.

The advantages of automation are mainly savings in space, savings on labor costs, 24/7 availability (it is not always easy to find unskilled personnel willing to do warehouse work), and savings on other operational costs, such as heating and lighting. Furthermore, robotic technologies provide scalability and throughput flexibility, which is essential in e-commerce environments where the demand variability is high. Automation of storage and order picking requires considerable scale and a long-term vision, as the investments can be earned back only in the medium and longer term. Therefore, it is crucial to develop tools to help decision makers find the correct solutions for their warehouses.

— Robotized and Automated Warehouse Systems: Review and Recent Developments (Azadeh et al. 2019, conclusion, pp. 939–940).

These are literature-level advantages and investment caveats. They are not guarantees that every installation will save space or labor, scale efficiently, or reach payback on the same timetable.

References

6 River Systems. 2020. “6 River Systems Solutions Brochure.” https://6river.com/wp-content/uploads/2020/03/6RS-SOLUTIONS-BROCHURE_03122020.pdf.

6 River Systems. n.d. “Meet Chuck.” Accessed September 3, 2026. https://6river.com/meet-chuck/.

AutoStore. 2025. “Warehouse Automation Boosts Hayabusa.” April 30. https://www.autostoresystem.com/cases/warehouse-automation-a-boon-to-japans-hayabusa.

AutoStore. n.d. “AutoStore Solution FAQ Resource.” Accessed September 5, 2026. https://www.autostoresystem.com/solution-faqs.

Azadeh, Kaveh, René De Koster, and Debjit Roy. 2019. “Robotized and Automated Warehouse Systems: Review and Recent Developments.” Review. Transportation Science 53 (4): 917–45. https://doi.org/10.1287/trsc.2018.0873.

CEVA Logistics, and Exotec. 2023. “CEVA Logistics Signs New Global Partnership with Exotec.” July. https://www.exotec.com/app/uploads/2023/07/CEVA-Logistics-x-Exotec-PR-July-2023.pdf.

DSV. 2020. “DSV Improves Warehouse Operations with Drone System.” November 17. https://www.dsv.com/en/about-dsv/press/news/com/2020/11/dsv-improves-warehouse-operations-with-drone-system.

Fizyr. n.d. “Palletizing and Depalletizing.” Accessed September 3, 2026. https://www.fizyr.com/cases/palletizing-depalletizing.

Fraunhofer Institute for Material Flow and Logistics. n.d. “RackRacer.” Accessed September 3, 2026. https://www.iml.fraunhofer.de/en/fields_of_activity/material-flow-systems/machines_and_facilities/developments/rackracer.html.

Manzini, Riccardo, Riccardo Accorsi, Giulia Baruffaldi, Teresa Cennerazzo, and Mauro Gamberi. 2016. “Travel Time Models for Deep-Lane Unit-Load Autonomous Vehicle Storage and Retrieval System (AVS/RS).” International Journal of Production Research 54 (14): 4286–304. https://doi.org/10.1080/00207543.2016.1144241.

“Multi-Deep Shuttle Storage System ‘D3’ | Automated Storage ＆ Retrieval System (AS/RS) | Products | Intralogistics | Solutions.” n.d. DAIFUKU. Accessed June 21, 2023. https://www.daifuku.com/solution/intralogistics/products/automated-warehouse/multideep-asrs/.

OPEX Corporation. n.d. “Perfect Pick.” Accessed September 3, 2026. https://www.opex.com/product/perfect-pick/.

Toyota Material Handling Nederland B.V. n.d. “AutoStore: De Toekomst van Magazijnautomatisering Met Toyota Material Handling.” Accessed September 5, 2026. https://toyota-forklifts.nl/magazijn-automatisering/automatisering-logistiek/autostore/.

U.S. Patent and Trademark Office, Patent Trial and Appeal Board. 2022. “Petitioner’s Exhibit List, IPR2022-00673, u.s. Patent 10,961,051, AS-1013.” https://ptacts.uspto.gov/ptacts/public-informations/petitions/1549264/download-documents?artifactId=a0Nq9ux8HaZGa704iDnkxDOPFhsbLy7dItu86cczHSCa2pozaJTJHNc.

Vanderlande. n.d. “ADAPTO — Systems.” Accessed September 3, 2026. https://www.vanderlande.com/warehousing-vanderlande/systems/.

Zaerpour, Nima, Yugang Yu, and René de Koster. 2017. “Small Is Beautiful: A Framework for Evaluating and Optimizing Live-Cube Compact Storage Systems.” Transportation Science 51 (1): 34–51. https://doi.org/10.1287/trsc.2015.0586.