The Series C funding was spearheaded by SoftBank Group Corp., with involvement from Nvidia Corp., Macquarie Group Ltd., 1789 Capital and the private investment firm Bezos Expeditions, belonging to Jeff Bezos, stated co-founder and Chief Executive Officer Deepak Pathak.
The funding also witnessed engagement from several existing backers, notably Lightspeed Venture Partners, Felicis Ventures, Sequoia Capital and Coatue Management. Further contributors encompassed Samsung Electronics Co. , LG Electronics Inc. and Salesforce Inc. The startup, headquartered in Pittsburgh, Pennsylvania, has secured roughly $2 billion thus far, according to Pathak. The enterprise was valued at approximately $4.5 billion during June of the prior year.
Established in 2023, Skild is an element of a burgeoning industry of startups that are garnering the interest of principal investors intrigued by the prospective of AI for robotics. In contrast to firms concentrated on particular functions, Skild is engineering artificial intelligence to formulate a comprehensive “intellect” for robots. The objective isn’t to streamline a solitary procedure, but to construct a framework capable of adapting to diverse settings and duties, assimilating knowledge similarly to humans: via observation and application. “There isn’t an Internet for robots,” Pathak articulated. “You are unable to devise an intellect for robots devoid of information, hence from the outset we prioritized the progression of a universal intellect for any machine, any task, any circumstance.”
Skild Brain software is operable on conventional GPUs without the necessity of a bespoke architecture. The framework is trained on extensive archives of human-created videos and practical simulations. Errors and actions generate novel information within the actual realm, yielding feedback that refines proficiency. The framework consolidates internal signals, such as joint movement and exertion, with external perceptual information, encompassing visuals, thereby enabling the robot to concurrently comprehend itself and its surroundings.
“Should one of the arms malfunction, the robot proceeds to execute the function,” remarked Abhinav Gupta, president and co-founder of Skild. “Should a leg falter, it doesn’t topple. This furnishes a novel degree of security that robots previously lacked.”
The recent investment will facilitate the broadening of Skilld Brain’s accessibility, enhance training, and implement additional robots across a spectrum of environments. The software exhibits compatibility with an array of robot variants—spanning from quadrupeds to manipulators and arms—in hospitals, residences, warehouses, and construction locales. It further aids humanoid robots in undertaking intricate physical actions: leaping over impediments, seizing objects, and executing tasks such as loading dishwashers, tidying, and preparing eggs.
Skild asserts to have expanded from zero to tens of millions of dollars in income within a handful of months during 2025. The startup is collaborating with over eight clientele, whose identities remain undisclosed. One client dispatched a service robot to LaGuardia Airport in New York to gauge air quality. LG CNS Co. is cooperating with Skild on humanoid resolutions.
The corporation’s escalating valuation and fresh investment round mirror an overarching inclination: prominent tech enterprises are progressively investing in startups instead of executing substantial acquisitions. Samsung perceives the investment in Skild as a means of preserving visibility behind the startup and its aptitude absent a significant commitment, and Nvidia underpins other entities within the domain, including Figure AI and Serve Robotics. Skild isn’t the sole startup endeavoring to render robots more astute and adaptable. Physical Intelligence, similarly backed by Samsung, was appraised at $5.6 billion in the preceding year and is operating in a comparable direction.
The sector confronts persistent engineering obstacles, notably for humanoid robots intended for the consumer sphere: batteries, maneuverability, and safeguarding impede widespread adoption. As an illustration, Tesla persists in encountering complications in introducing its Optimus robot to the marketplace.
As per Craig Le Clair, vice president and principal analyst at Forrester Research, the robotics sector is presently undergoing a “GPT juncture,” transitioning from single-task mechanisms to general-purpose robots. Emerging “vision-language-action” models are learning to maneuver by scrutinizing millions of hours of video, facilitating the mastery of fundamental motor proficiencies such as ambulation and folding garments.
Skild employs in excess of 100 individuals, numerous of whom have previously been employed at Meta, Tesla, Nvidia, Amazon, and Google. The concept for the startup materialized in early 2022, antecedent to the unveiling of ChatGPT. The co-founders, who were engaged at Meta’s robotics laboratory and served as professors at Carnegie Mellon’s Robotics Institute, discerned the constraints of academic endeavors and resolved to establish a company capable of scaling solutions for tangible challenges.
In February 2024, the team procured a commercially procurable humanoid robot from Unitree, installed its software, and accomplished noteworthy outcomes within a singular day. Skild emerged from stealth mode during the summer of 2024 and in the ensuing year showcased its inaugural public videos of its frameworks in operation. At Nvidia’s GTC in October, CEO Jensen Huang proclaimed that Skild’s software would be utilized to robotize a novel GPU factory in Houston.
Skild additionally underpins a more encompassing initiative for sophisticated automation within the US, postulating that robots will generate employment opportunities and populate over 1 million perilous, repetitive, or unwanted positions.
“Ultimately, robots will possess the capability to execute nearly any function a human is capable of, and that constitutes a favorable, not an adverse, development,” Pathak conveyed. “We shall possess ample time to prepare for that. It will not transpire instantaneously, nor will it materialize within a handful of years.”
Source: Bloomberg