MERPHI Robotics

Here it is: Bimanual Wheel-Based Humanoids Prenote: The purpose of this infographic is to educate the public and the industry about the progress of the robotics sector. There is no paid affiliation with any of the companies represented. This infographic explores one of the fastest-growing categories in robotics in Q1 2026: bimanual hybrid/wheel-based humanoids. These robots are often on a mobile base, equipped with two arms and multi-finger grippers, an HRI interface or “head,” and designed to perform a variety of tasks across multiple sectors. While they may not be as versatile as bipedal humanoids, they are certainly more advanced and versatile than many other categories of robots. For more context, please refer to the robotics categorization linked in the comments. We handpicked some of the most notable robots in this field from recent years up to June 2026. The goal was not to showcase every single robot, but to highlight the most impactful ones. For some robots, the specifications are incomplete. This is either because the company has not published a full spec sheet, or because the system has multiple variants and can be customized for different customers. Note: We have been consistently building robotics-related infographics for the past 3.5 years. Although this is not MERPHI’s main mission, and all of our infographics are open to the public and free of charge, we would appreciate it if you reference our work when republishing. A new version will be released in approximately five months, where we will track the progress of this category and add or exclude robots accordingly. Please feel free to DM me or Filip for the high-quality version until we publish it on the website, where you will be able to download the printable PDF version. Cheers #humanoids #humanoid_posters #robotics #hri #robot_design #robot_market #physical_AI #humanoid_infographics

Here it is: Bimanual Wheel-Based Humanoids Prenote: The purpose of this infographic is to educate the public and the industry about the progress of the robotics sector. There is no paid affiliation with any of the companies represented. This infographic explores one of the fastest-growing categories in robotics in Q1 2026: bimanual hybrid/wheel-based humanoids. These robots are often on a mobile base, equipped with two arms and multi-finger grippers, an HRI interface or “head,” and designed to perform a variety of tasks across multiple sectors. While they may not be as versatile as bipedal humanoids, they are certainly more advanced and versatile than many other categories of robots. For more context, please refer to the robotics categorization linked in the comments. We handpicked some of the most notable robots in this field from recent years up to June 2026. The goal was not to showcase every single robot, but to highlight the most impactful ones. For some robots, the specifications are incomplete. This is either because the company has not published a full spec sheet, or because the system has multiple variants and can be customized for different customers. Note: We have been consistently building robotics-related infographics for the past 3.5 years. Although this is not MERPHI’s main mission, and all of our infographics are open to the public and free of charge, we would appreciate it if you reference our work when republishing. A new version will be released in approximately five months, where we will track the progress of this category and add or exclude robots accordingly. Please feel free to DM me or Filip for the high-quality version until we publish it on the website, where you will be able to download the printable PDF version. Cheers #humanoids #humanoid_posters #robotics #hri #robot_design #robot_market #physical_AI #humanoid_infographics

This new infographic is one of our most robot-rich posters so far, and it will be unveiled at the Humanoids Summit in Tokyo on May 29. If you are joining us there, you can attend my keynote on Friday. If not, stay tuned. We are excited about this one because this category has been one of the fastest-growing areas in #robotics and #humanoids over the past five years. Our selection criteria were based on the following: bimanual arms, end-effector inclusion, an HRI or head module, a mobile base, active or passive, and a proven operational track record within the past three years in a clearly defined focus industry. Wait for the big reveal in a few days :) Filip Sperr Philip Justinus Berlin

Finally, I’m sharing a personal open-source project I’ve been working on around #humanoid #robot design. Over the years, I’ve repeatedly seen the same pattern: early mechanical decisions are often based on human proportions, and only later adjusted once simulation, control, or task constraints come in. At that point, changing geometry becomes expensive and inconsistent across teams, especially in #industrial_design. GulaMannen is an attempt to address that. It’s a modular humanoid reference platform intended for simulation-first exploration. Instead of starting from scratch or assembling a generic placeholder, teams can use a set of predefined configurations and swap key parameters such as joint architecture, base type, and proportions to test different directions early. The current version includes six configurations: - a full biped with serial joints - a biped with parallel wrist and ankle - several hybrid setups (pedestal, vertical actuator, rail-based, single-leg) They all share a common upper body and a consistent actuator logic, so comparisons between setups are easier. I also added a small set of reference environments (desk, shelves, pallet/box setups). The idea is to evaluate reach, height, and task compatibility directly, not just kinematics in isolation. The CAD is hosted on Onshape and is fully accessible. URDF is available for some configurations and can be extended. The system is modular, so parts can be recombined or adjusted depending on the use case. This is still a work in progress. Mass properties and dynamics are not fully refined across all configurations, and simulation coverage can be improved. The intention is to open it up and let it evolve with input from others working on similar problems. If you’re working on humanoids and need a starting point for simulation or early design exploration, feel free to take a look and adapt it. Thanks to Ben Katz (MIT), Scott Walter, Ph.D., Connor Shannon, Stephen Morfey, Rob Knight, and others for early feedback. Special thanks to Kimate Richards, who is supporting URDF generation on demand. Links in the cmnt section.