Indoor Mobile Robots

Loading and unloading containers and loose-load trucks can involve some of the most physically demanding activities in logistics. Workers are exposed to harsh climate conditions and must repeatedly move heavy goods in confined spaces as quickly as possible to ensure downstream operations can continue uninterrupted.

More and more companies are exploring ways to automate these tasks. To unload pallets from a trailer, Fox Robotics has developed the Automated Trailer Unloading solution which allows one operator to control several robots simultaneously via a tablet. Boston Dynamics has developed the Stretch robot for a range of different warehouse tasks, but initially this indoor mobile robot will enable autonomous unloading of floor-loaded containers and trucks. Stretch consists of three components: a lightweight robotic arm that can lift cartons of up to 22.7 kg (50 lb), a relatively small mobile base designed to easily fit inside trailers and containers, and the perception arm that includes depth sensors and 2D cameras to help the robot identify cartons of different shapes and sizes and provide insights that help Stretch perceive its operational environment. Future models may be capable of loading, building up pallets, and depalletization. In January 2023, DHL Supply Chain became the first company to achieve commercial application of Boston Dynamics’ Stretch robot developed for carton unloading of trucks.

Today’s deployed solutions are typically for unloading uniform cartons or pallets. Loading and unloading of mixed cases are complex tasks and it will take further research and development before scalable fully autonomous solutions come to market. However, with advances in robotic software and computer vision capabilities, indoor mobile robots will soon achieve additional autonomous tasks in mobile piece manipulation.

In the unautomated warehouse environment, workers can sometimes be required to walk up to an average of 14.5 km (9 miles) each day. This puts an enormous physical burden on workers and at the same time presents an opportunity and a strong argument for the introduction of indoor mobile robots. Assisted picking robots can drive efficiency in the overall order fulfillment process by shortening the distance walked by humans and reducing the time between picks.

There are different types of order fulfillment applications including goods-to person solutions, and solutions that fall into a zone picking category. A classic goods-to-person AMR solution comes from Geek+ with robots able to pick up inventory shelves and transport them to the required picking or pack station where an employee prepares the order for fulfillment without having to walk anywhere.

In zone-picking applications, workers are assigned to operate in a specific predefined warehouse area. Locus Origin is a prime example of a direct zone-picking solution, where totes are placed on robots that travel to pickers in various work zones and then visually indicate the closest pick option for each tote via a screen. At DHL, we have recently expanded on our partnership with Locus to encompass a total deployment of 5,000 Origin bots across warehouses and distribution centers globally and we have celebrated our 500 millionth pick by Origin bots globally.

Using AMRs to support order fulfillment allows workers to focus on the accuracy of their picking and significantly diminishes the distance they must cover each day and the required travel time.

Significant progress in the development of perception technology and AMR capabilities has encouraged further investment in automated solutions to support facility functions like cleaning, security, and maintenance and repair operations (MRO).

Since 2020, the global warehousing market has seen significant growth, driven largely by the surge in e-commerce and increasing demand for more sophisticated logistics. Such growth in the footprint of warehousing operations has driven additional investment in and deployment of intelligent cleaning robots. Solutions like the Karcher KIRA bot have enabled smartification of the cleaning process by deploying cloud-based artificial intelligence (AI); KIRA can learn and improve its cleaning performance over time. Additionally, smart detection systems allow for safe navigation of the bots around the people and obstacles that are present in the warehousing environment.

With increasing sprawl and the resulting higher requirements for security infrastructure, today’s warehouses are also at risk of theft. Companies can integrate autonomous security robots (ASRs) to better monitor and control these facilities. Use cases range from automatic license plate recognition to the detection of blacklisted mobile devices or suspicious devices and video surveillance.

The four-legged mobile robot Spot, from Boston Dynamics, can be used as an ASR, equipped with additional cameras and night-vision technology. Designed to walk stairs and uneven terrain, this robot can live monitor different levels in a warehouse or operations facility and can be operated remotely via a tablet or follow pre-mapped routes.

The use of cleaning bots and ASRs in warehouse cleaning and security functions remains limited but, in future, it is likely that indoor mobile robots will play a significant role in intelligent facility support infrastructure.

Warehouse space utilization is becoming increasingly important. For example, the overall cost of warehousing properties across 52 international markets rose by an average of 8.4% (12 months to June 2022), largely due to heightened demand for warehouse space, a consequence of the pandemic, and e-commerce expansion.

One way of boosting storage density is to introduce an automated storage and retrieval system (ASRS). These computer-controlled robotic storage systems can be used for autonomous handling, storage, and retrieval of totes or shelves in a warehouse. ASRS solutions, such as the cube-based system from Norwegian company AutoStore, have many advantages over traditional picking solutions. AutoStore’s solution offers a high-density system reducing the required space by a factor of up to 4 and can speed up manual picking by up to 5 times compared to conventional picking.

Although the initial investment for such solutions is relatively high, this can be economically viable, especially when handling large volumes of stock-keeping units (SKUs) of slow- and medium-velocity goods, optimizing the throughput and operational efficiencies of the system. DHL Supply Chain is currently involved in nine AutoStore deployments across Singapore, the US, and Germany, with four more in planning. Through this partnership, a fleet of over 1,000 robots has significantly boosted operational efficiency and throughput in DHL’s fulfillment warehouses.

In recent years, there’s been steady development and commercialization of humanoid robots – machines designed to resemble humans and perform similar tasks using human-like movements and interactions. In logistics operations, these robots could relieve personnel from monotonous tasks that cannot be performed by conventional stationary robotics, such as loading and unloading pallets and containers.

While humanoid robots may appear to provide an ideal solution, it may not make sense to construct legs if wheels are the swifter option for moving around the warehouse. Similarly, there may be no need to equip the machine with two arms if one is sufficient to do the job. Rather than recreating the human body, it may be better to design robots according to their tasks and for optimal interoperability.

There is significant development potential for humanoid robots, particularly in achieving a completely autonomous and non-externally controlled flow of activities. Companies like Boston Dynamics with its Atlas robot, the humanoid-specialized company Figure AI, and the Chinese company Unitree, which offers its G1 at a relatively low price point, are making significant progress in this technology field. Nevertheless, in future, even the most advanced robots will not completely replace humans in logistics.