RealMan Robotics Open-Sources RealSource Multimodal Robot Dataset

RealMan Robotics has released RealSource, an open multimodal dataset built from real-world robot interactions. The dataset aims to address data alignment gaps in embodied AI and robotics research.

By RB Team Published: Updated:

RealMan Intelligent Technology Co. has announced the open-source release of RealSource, a multimodal robot dataset collected entirely from real-world environments. The dataset was generated at the company’s Beijing Humanoid Robot Data Training Center and is designed to address persistent shortages in fully aligned, high-quality data for embodied AI and robotic manipulation research.

RealSource spans 10 real-world scenarios, including smart homes, eldercare, retail, agriculture, catering, and automotive assembly. Data was captured using three different robotic platforms performing tasks such as folding laundry, opening appliances, and sorting materials. The dataset integrates synchronized RGB and depth vision, joint states, force sensing, action commands, and precise timestamps, covering the full perception-to-execution pipeline.

The release reflects growing emphasis on realistic, multimodal datasets to improve generalization in robotics. By prioritizing sensor alignment, noise resistance, and repeatability across environments, RealMan is positioning RealSource as infrastructure for researchers developing robust, real-world robotic systems.

News, Robots & Robotics

GS25 Becomes First Convenience Store Chain to Sell Humanoid Robots in South Korea

South Korean convenience store chain GS25 has begun selling humanoid robots, quadrupeds, and AI social robots as part of its Family Month curated lineup. The retailer is offering 11 robot products, marking the first time humanoid robots are being sold through the convenience store channel.

By Rachel Whitman Published:

South Korean convenience store chain GS25, operated by GS Retail, has begun selling humanoid robots through its retail channel as part of a curated lineup of more than 130 Family Month items. The retailer is offering 11 robot products, including the AI social robot Liku, a G1 humanoid robot, an Air quadruped walking robot, and Ailico robot keyrings. According to GS Retail, this marks the first time humanoid robots are being sold through the convenience store format in the country.

Liku, positioned as the flagship product, is a Korean-developed AI social robot equipped with conversation and emotion-expression capabilities, with target use cases in children’s education and elderly care. The broader robot lineup ranges from full-sized humanoid and quadruped platforms to smaller form factors such as keyring devices designed for portable, low-stakes interaction. The robots are being offered alongside non-technology items including pure gold and silver bars, Korean beef, seafood, and fruits as part of the chain’s Family Month promotion.

The move signals a shift in how consumer-facing robotics products are reaching the mass market in South Korea. Rather than being confined to specialty retailers, electronics chains, or direct-from-manufacturer sales, robots are now being positioned alongside everyday consumer goods. This approach treats robots as gift or lifestyle items, lowering the perceived barrier for first-time buyers and broadening the distribution surface for early-stage humanoid and social robot products.

Whether the convenience store channel becomes a meaningful sales pathway for humanoid robots will depend on price points, after-sales support, and consumer interest beyond the initial promotional window. Still, the listing reflects the maturation of South Korea’s domestic robotics ecosystem and the willingness of major retailers to test embodied AI products in formats traditionally reserved for fast-moving consumer goods.

News, Robots & Robotics

Toyota Woven City Begins Real-World Testing of Robotics and Autonomous Mobility Near Mount Fuji

Toyota’s Woven City in Japan has moved from concept to operational testbed, with around 100 early residents already living alongside delivery robots, autonomous vehicles, and AI systems. The site near Mount Fuji is positioned as a long-term proving ground for mobility and human-machine interaction.

By Daniel Krauss Published:

Toyota’s Woven City, an experimental urban development near Mount Fuji, has transitioned from a concept project into an active testbed for robotics, artificial intelligence, and autonomous mobility. The first phase covers roughly 47,000 square meters and is set to expand to about 294,000 square meters at full build-out. Around 100 early residents, referred to as “weavers”, are already living on site, with long-term plans for up to 2,000 inhabitants. The project carries an estimated long-term cost of around $10 billion.

The development is structured as a functioning urban environment rather than a demonstration zone. Underground passageways handle logistics and mobility systems, while above-ground spaces host trials involving home robots, autonomous transport prototypes, and AI safety infrastructure. Concepts such as flying taxis are being explored in simulation. Residents participate directly in evaluation cycles, providing feedback on usability and human-machine interaction that feeds back into product development.

The closed-but-populated model is intended to address a regulatory gap that limits large-scale autonomous vehicle and robotics testing on standard public roads. By controlling the operating environment while maintaining real pedestrian and resident behavior, Toyota and its partners aim to collect data on mobility patterns, safety incidents, and everyday usage that would be difficult to gather on conventional streets. Developers have indicated this evidence is intended to inform future regulatory frameworks for autonomous transport.

Woven City reflects a broader industry shift toward integrated, real-world environments for validating embodied AI and autonomous systems, moving beyond isolated lab tests and individual pilot deployments. Similar approaches are being explored elsewhere, but the scale and corporate backing of Toyota’s project make it one of the most substantial efforts to combine robotics, mobility, and urban infrastructure in a single operating site. Its long-term influence will depend on whether the data and design lessons generated translate into deployable systems beyond the controlled perimeter.

Artificial Intelligence (AI), News, Robots & Robotics, Science & Tech

SquareMind Raises $18M to Scale Swan Robotic Dermatology Imaging Platform

French medical robotics company SquareMind has raised $18 million to commercialize Swan, a robotic platform that performs automated full-body dermoscopic skin imaging. The round was led by Sonder Capital, the venture fund co-founded by Intuitive Surgical’s Fred Moll.

By Laura Bennett | Edited by Kseniia Klichova Published: Updated:

Medical robotics company SquareMind has raised $18 million to commercialize Swan, a robotic platform that performs automated full-body dermoscopic skin imaging for dermatology practices. The round was led by Sonder Capital, the venture fund co-founded by Intuitive Surgical founder Fred Moll, with participation from the Deeptech 2030 Fund managed by Bpifrance on behalf of the French government, Adamed Technology, Calm/Storm Ventures, Teampact Ventures, and several entrepreneurs. The capital will fund commercial, engineering, and customer support hiring ahead of Swan’s launch in the United States and Europe.

Swan acts as an augmented dermatoscope, capturing standardized images of the entire skin surface at a resolution typically reserved for close-up examination of individual moles. SquareMind says it is the first robot to perform this kind of automated full-body dermoscopic capture. During a session, the patient stands in a private exam room while a robotic arm moves around them, guided by visual and audio prompts, with image acquisition completed in minutes and without physical contact. The platform is paired with AI-based review software that helps track new or changing lesions over time, while clinical judgment remains with the physician.

The company is targeting a structural bottleneck in dermatology. Skin screening is the highest-volume procedure in the specialty, and waitlists in many markets now stretch into months as an aging population drives demand. SquareMind notes that roughly 80 percent of melanomas appear as new lesions rather than changes to existing ones, making consistent full-body documentation important for early detection but difficult to achieve under typical appointment time constraints.

The investment reflects continued capital flow into clinical robotics platforms designed to automate high-volume, documentation-heavy procedures rather than complex surgical interventions. Whether Swan can establish itself as a standard tool will depend on integration with clinical workflows, reimbursement pathways, and the ability of its imaging and AI-assisted review pipeline to produce outcomes that justify the equipment footprint in dermatology practices.

Business & Markets, News, Robots & Robotics

Canadian Robotics Council Forms Capital Committee to Boost Domestic Robotics Investment

The Canadian Robotics Council has created a Capital Committee to connect robotics startups with banks, venture firms, and specialized financing. The initiative aims to increase domestic investment in Canadian robotics and physical AI companies.

By Daniel Krauss Published: Updated:

The Canadian Robotics Council has established a Capital Committee designed to increase investment in Canada’s robotics sector. The group brings together representatives from major financial and venture organizations, including BDC Capital, Garage Capital, Inovia Capital, RBC Dominion Securities, Two Small Fish Ventures, and Version One Ventures. According to BetaKit, the committee was created to strengthen links between capital providers and robotics companies that often need significant funding to scale hardware, manufacturing, and deployment operations.

The committee will focus on three areas: expanding funding for robot makers and automation adopters, giving investors better technical frameworks for evaluating robotics startups, and connecting entrepreneurs with supply chains, early customers, and specialized financing. The Canadian Robotics Council argues that Canada has strong robotics research and early-stage innovation, but scaling companies requires deeper investor understanding of hardware risk, deployment timelines, and technical due diligence. Its founding members have already backed Canadian robotics companies including Avidbots, Clearpath Robotics, Haply Robotics, Kindred Systems, and Waabi.

The move reflects a broader concern that Canada could underinvest in physical AI while focusing more heavily on software-based artificial intelligence. Robotics companies typically require more capital than pure software startups, but they can also create stronger industrial spillovers through manufacturing, automation adoption, and domestic supply chains. By formalizing investor engagement, the council is positioning robotics as a strategic infrastructure sector rather than a narrow hardware category.

News, Policy & Regulation, Robots & Robotics

Cognibotics Wins $1.7M Order for Robot Calibration Technology

Cognibotics has secured a $1.7 million order for its CogniCal robot calibration system from a major industrial robot manufacturer. The deal highlights growing demand for precision calibration in large-scale automation.

By Rachel Whitman Published:

Cognibotics has received an order worth approximately $1.7 million for its CogniCal robot calibration technology from a major global industrial robot manufacturer. The order expands the customer’s use of CogniCal within robot production processes and reflects increasing emphasis on precision calibration as automation systems scale.

CogniCal improves robot accuracy by compensating for geometric deviations, structural elasticity, and payload effects that can vary between individual machines. By aligning robot motion more precisely, the system reduces performance variance across fleets and shortens commissioning time for integrators and end users deploying automation systems.

The demand for calibration technologies is rising as robotics systems incorporate perception and AI-driven workflows. Inconsistent robot geometry can introduce noise into training data and limit the transferability of learned behaviors. Calibration layers such as CogniCal are therefore emerging as core infrastructure for scalable industrial robotics and AI-enabled automation.

News, Robots & Robotics