Four companies with humanoid robots either in commercial deployment or late-stage trials will share a stage at RoboBusiness to discuss where the technology stands in 2026. Agility Robotics, Apptronik, Persona AI, and PSYONIC will each send representatives to a panel focused on the current state of humanoid development, deployment economics, and the timeline for scaled adoption. The panel format is notable because it places direct competitors in conversation with each other at a moment when customer skepticism about humanoid robots remains high despite significant venture investment in the category.
Agility Robotics operates the furthest along the commercialization curve among the four panelists. The company's Digit robot has been deployed in Amazon fulfillment centers since late 2024, with additional pilots running at Spanx and other logistics operations. Agility has raised over $330 million in total funding, most recently a $150 million Series C in 2025, and began production at a 10,000-unit-per-year manufacturing facility in Salem, Oregon earlier this year. Digit stands 5 feet 9 inches tall, weighs 141 pounds, and is designed specifically for tasks like tote moving and container handling in structured warehouse environments. The robot's narrow profile allows it to navigate spaces built for human workers without facility redesign, a selling point Agility emphasizes when pitching logistics customers wary of infrastructure costs. Apptronik, by contrast, has focused its Apollo humanoid on automotive and electronics manufacturing applications. The Austin-based company announced partnerships with Mercedes-Benz and NASA in 2024 and has emphasized modularity in Apollo's design, allowing the robot's torso, arms, and lower body to be reconfigured for different tasks. Apollo can lift up to 55 pounds and stands just under 5 feet 8 inches. Apptronik has raised approximately $75 million and is backed by investors including Fidelity and Oakhouse Partners. The company has not disclosed production volumes but indicated in early 2026 that it expects to ship "hundreds" of units by year-end.
Persona AI represents a different approach entirely. Rather than building hardware, the San Francisco-based startup develops foundation models specifically for humanoid robot control and navigation. Persona's software has been integrated with several third-party humanoid platforms, and the company positions itself as an enabler for hardware manufacturers that lack deep expertise in real-time decision-making and manipulation planning. Persona raised a $40 million Series A in late 2025 led by Lux Capital and counts Sanctuary AI and several unnamed Chinese manufacturers among its customers. The inclusion of Persona on the panel reflects growing recognition that humanoid success depends as much on software maturity as mechanical design. PSYONIC rounds out the group with a focus on bionic limbs rather than full-body humanoids, but the company's Ability Hand shares core technologies with humanoid manipulators, including touch feedback, adaptive grasp control, and lightweight actuation. PSYONIC has shipped over 400 hands to amputees and research institutions since 2022, making it the most commercially established among the four panelists by unit volume. The company's presence signals that the panel will address not just humanoid locomotion but also the manipulation and haptics challenges that every humanoid developer faces.
The broader context for this discussion is a humanoid robot market that has attracted more than $2 billion in venture funding since 2023 but has yet to demonstrate profitability at scale for any participant. Tesla's Optimus remains in internal testing with no confirmed third-party sales. Figure AI, despite raising $675 million at a $3.2 billion valuation in early 2025, has disclosed only pilot deployments with BMW. Boston Dynamics has positioned its Atlas robot as a research platform rather than a commercial product. Sanctuary AI's Phoenix has been tested in retail environments but not deployed at volume. Industry analysts point to a persistent gap between investor enthusiasm and customer willingness to pay for humanoid capabilities that more specialized automation can often deliver at lower cost and higher reliability. The RoboBusiness panel takes place against this backdrop of high expectations and limited proven business models. Questions about cycle time, mean time between failures, operator training requirements, and total cost of ownership remain largely unanswered in public forums. Whether the panelists will offer concrete performance data or stick to high-level narratives about long-term potential will determine whether the session delivers value to engineers and investors tasked with deciding where to allocate resources in 2027.
What to Watch: Monitor whether Agility discloses specific uptime or utilization metrics from its Amazon deployment, as any quantitative performance data would provide the first independent benchmark for humanoid reliability in logistics. Track Apptronik's production ramp in the second half of 2026 to see if the company meets its stated goal of hundreds of units shipped. Watch for customer announcements from Persona AI, particularly any deals with Western manufacturers that would validate the market for humanoid-specific control software outside China. Pay attention to any new partnerships PSYONIC announces, especially if they involve integrating Ability Hand technology into full humanoid platforms from other developers.
