How Humanoid Robots Could Close Gaps in Packaging Automation

What sounded like science fiction just a few years ago is now a reality: humanoid robots are picking up objects, transporting containers and taking on their first tasks in industrial processes.

The development of humanoid robots for packaging builds on the progress made by pick-and-place robots and cobots in recent years. Using machine learning and intelligent sensors, machinery suppliers have enabled new applications and closed gaps in packaging line automation. However, frequently changing manual tasks and uncommon applications remain a challenge. This is where humanoid robots are expected to play a role: they could combine different handling operations and take on tasks at workstations previously reserved for people. Fraunhofer IML therefore describes them as flexible, multipurpose robots that combine perception, grasping, movement and interaction.

Early Applications in Packaging and Logistics

Humanoid robots could create new opportunities for packaging automation, particularly where small batch sizes and frequent product changes require flexible processes. Several technology companies are developing models specifically for these tasks.

In 2025, US developer Figure AI presented a logistics application in which humanoid robots transfer parcels between conveyor belts and orient their labels for scanning. The variety of items presents a particular challenge: boxes, bags and envelopes differ in size, weight, shape and rigidity. The robot must detect moving objects, select a suitable grip and rotate each item as it handles it. It is controlled by Figure’s proprietary AI model, Helix. According to the company, learned behaviour can be transferred to other robots. This could allow users to automate repetitive handling tasks despite varying parcel formats.

US robotics developer Apptronik is also targeting packaging tasks. Its humanoid robot Apollo is intended to place picked items into shipping boxes and transfer completed parcels to conveyor belts, according to its packout product page. Apptronik is also working with logistics provider GXO to trial specific warehouse applications. A press release on the partnership lists transporting boxes and containers, picking individual items and scanning barcodes among the development objectives.

Bosch is also involved in developing humanoid robots. From 2027, the group plans to manufacture robots for British technology company Humanoid at its Bühl plant. Humanoid lists goods handling and “Picking & Packing” in fulfilment centres and e-commerce warehouses as potential applications.

Whether the robots manufactured at Bosch’s Bühl plant will actually be used in packaging processes remains open. However, initial industrial trials offer indications of the capabilities required. During a joint proof of concept with Schaeffler in Schweinfurt, an HMND 01 retrieved metal bearing rings from a container. The first deployment phase, scheduled to begin in December 2026 in Herzogenaurach, is set to include box handling, as Reuters reports.

Siemens is also collaborating with Humanoid. At its electronics factory in Erlangen, the company tested the wheeled HMND 01 Alpha in autonomously picking up, transporting and placing containers.

These trials put the robot’s capabilities to the test in areas that are also essential to packaging processes.

Flexibly Combining Multiple Tasks

Packaging often involves a complex sequence of manual tasks: identifying, picking up, transporting and placing items, opening or filling packaging, and digitally confirming the operation. As the Fraunhofer Institute for Material Flow and Logistics IML explains in its study “Automation on two legs? Humanoid robots in logistics”, changes in objects, locations and process conditions make end-to-end automation difficult. The key is therefore not simply the individual grasping operation, but the flexible combination of several process steps.

The companies surveyed for the study therefore expect humanoid robots to offer flexible deployment options rather than perform a single, particularly difficult task. Depending on requirements, such a system could pick orders, fill boxes, supply packaging materials or transport objects between workstations.

This versatility could be particularly useful when workloads fluctuate, orders peak seasonally or staff are absent at short notice. However, deployment will only be economically viable if the robot can be integrated into existing processes without extensive modifications and quickly adapted to new tasks.

What Humanoid Robots Still Need to Reach Market Readiness

In packaging processes, these systems must handle lightweight, delicate, irregularly shaped or flexible items as reliably as heavy products and boxes. The industry’s long-term goal is therefore to achieve flexibility comparable to that of humans. According to Fraunhofer IML, however, this goal remains beyond the capabilities of current technology. The main shortcomings preventing widespread operational use are grasping precision, dexterity and reliable handling of a broad variety of objects.

Although 65 percent of the companies surveyed already use robotic solutions, humanoid systems remain exclusively in pilot projects or early trials. At the time of the market analysis, none of the suppliers examined could point to established operational applications in logistics.

Nevertheless, expectations are high: more than 80 percent of the companies surveyed expect humanoid robotics to be used in productive operations within the next ten years, with around 40 percent anticipating this within the next four years. At the same time, 43 percent of companies have no concrete deployment plans.

The future of humanoid robots in the packaging industry will therefore be determined in everyday operations. Only when they reliably handle changing items, flexibly combine multiple process steps and operate economically can individual pilot projects lead to widespread adoption.