Rhoda AI raises $450 million to train robots using internet videos at $1.7 billion valuation

AI Robotics Startup Rhoda Valued at $1.7 Billion in New Funding

Rhoda plans to make its own hardware and license its model to customers, with potential for a new robots-as-a-service business model where clients would rent the hardware and the software to meet their needs. A new AI startup is betting that it has found a powerful data source to teach robots to do more valuable work: online videos. Rhoda AI has raised $450 million in funding to develop and deploy a new artificial intelligence model trained on millions of publicly available internet videos, with the goal of directing robots to perform a variety of industrial tasks, including in unfamiliar conditions. The round was led by Premji Invest and values the startup at $1.7 billion, including the money raised. Currently, AI models for robotics are typically built using data from teleoperation, where humans remotely control the robot's motion. That limits the amount of data available for robots to adapt to the wide range of situations found outside a lab. Rhoda's team is betting that online videos can help fill in the gaps, along with a smaller amount of robot telemetry. "In the case of teleoperation, if the phone orientation changes, that might be enough to cause the model to fail," said Chief Executive Officer Jagdeep Singh, who previously founded and ran QuantumScape Corp., the solid-state battery maker. "Whereas in our case the model, it's seen so many other examples of objects that are at different orientations, it's able to generalize. Rhoda is one of a growing number of startups drawing significant interest from investors for the promise of using advances in AI to make robots better at navigating the world around them. Other backers in the new funding round include Khosla Ventures, Singapore's state-owned investor Temasek Holdings Pte and venture capitalist John Doerr. Vinod Khosla, who has invested in some of Singh's other companies, let Singh incubate Rhoda inside his namesake venture firm. Khosla Ventures also helped Singh recruit his co-founders. "The real world is messy, complex and being able to actually work on production lines is much, much harder than doing the demo," said Khosla, noting that he believes Rhoda has the right technology to do real work. Rhoda calls its approach a Direct Video Action system and claims it lets things like robotic arms deal with conditions that are different than the AI model expects. The company said it has run a successful test of its model using off-the-shelf parts in the factory of a leading automotive firm, and plans to license it to customers. The startup is also making humanoid robots, Singh said. Rhoda plans to make its own hardware in order to ensure it's capable of handling useful real-world work, as well as to collect data from its robots to further improve its model. Sandesh Patnam, managing partner at Premji, sees potential for a new robots-as-a-service business model, where clients would rent the hardware and the software to meet their needs. He said such a technology offering is increasingly important as the US tries to onshore more manufacturing. "If you want to really make the claim that we want to bring sophisticated high-end manufacturing back into the United States, we need this," he said.

Rhoda AI raises $450 million at $1.7 billion valuation, unveils robot intelligence platform

March 10 (Reuters) - Rhoda AI on Tuesday said it has raised $450 million in a Series A funding round that values the company at $1.7 billion and unveiled a robot intelligence system it says can handle the unpredictability of industrial environments. Breakthroughs in artificial intelligence models that help robots understand language, interpret visual information and predict how the physical world behaves, combined with growing ⁠investment from major tech and robotics companies, are expected to drive robotics adoption. Newer compact high-performance processors are equipping robots for real-time perception and broad-skill operation. That momentum is fueling a race in humanoid robots specifically, led by Tesla (TSLA.O), opens new tab, Figure AI, Unitree, Agility Robotics, and dozens of Chinese startups. However, industry experts caution that reliability, safety certification and cost will remain key hurdles for large-scale commercial deployment of general-purpose robots. Rhoda's new robot intelligence platform, FutureVision, works by first studying hundreds of millions of internet videos to learn how ⁠objects move and how the physical world behaves. It then uses that knowledge to constantly anticipate what is about to happen around it and translate those predictions into physical movements, a cycle it repeats dozens of times per second. The approach targets a longstanding ⁠problem in robotics: most machines perform well in controlled, predictable environments but struggle when something unexpected happens. The company expects to eventually license FutureVision to companies running robotic hardware and ⁠software platforms. Rhoda AI, which emerged from stealth on Tuesday, said its platform is designed to integrate with a wide range of robotic hardware, allowing manufacturers ⁠and logistics operators to deploy intelligent robots without rebuilding existing systems. Its funding round drew backing from Khosla Ventures, Temasek, Mayfield, Premji Invest and Capricorn Investment Group, among others. Reporting by Akash Sriram in Bengaluru; Editing by Sahal Muhammed Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Technology * ADAS, AV & Safety * Manufacturing * Sustainable & EV Supply Chain

Rhoda AI raises $450M to build foundational robotics models that learn from internet videos - SiliconANGLE

Rhoda AI raises $450M to build foundational robotics models that learn from internet videos Palo Alto-based Rhoda AI, a developer of foundational artificial intelligence robotics models, said today it raid $450 million in Series A funding to train intelligent robots using publicly available internet videos. Premji Invest led the round, bringing the valuation of the startup to $1.7 billion. The internet has been a prolific ground for providing examples of human activity. It's a perfect place to get candid presentations of almost every task, from picking up items of every kind to packing boxes, sorting objects and completing various manual jobs. Currently, most robots are trained through a process called teleoperation. This requires specialized remote-control equipment, including gloves and external sensors that collect positional data that can be translated into movements. Some systems can also work from numerous videos of actors taking the same action over and over. This limits the total amount of training data that can be input into robotics models. To handle this, most robotics companies extrapolate from these limited data sets by generating synthetic datasets. However, with the vast amount of data already available on the internet of hundreds of thousands to millions of humans already doing everyday jobs in natural environments, researchers have ready-made datasets. Rhoda's team said they believe that online videos will bridge the gap between limited data for orienting tasks by supplementing the small amount of robot telemetry with a broader array of natural human movements. "In the case of teleoperation, if the phone orientation changes, that might be enough to cause the model to fail," Chief Executive Jagdeep Singh told Bloomberg. "Whereas in our case, the model it's seen so many other examples of objects that are at different orientations, it's able to generalize." Teleoperation can give a robot a framework to understand the basics of a task, but it cannot give a robot the wherewithal to adapt. That's where online videos come in to provide the context, generalization, different orientations, failure states, varied approaches and edge cases. Something that's difficult to reproduce even with the "fuzzy logic" of synthetic generation. The company calls the approach Direct Video Action and said it allows setups such as robot arms to deal with edge conditions better than current AI training models. The company added that it has run successful tests using off-the-shelf parts with an automotive firm and has plans to license its AI model to customers. Singh said that Rhoda plans to build humanoid-style robots and has plans to make its own hardware to ensure quality for real-world work. Other backers joining the round included Khosla Ventures, Singapore's state-owned investor Temasek Holdings Pte and venture capitalist John Doerr. Rhoda joins a number of companies drawing investor attention for a growing advancement called Physical AI, where AI bridges the gap between the physical and digital worlds, allowing machines to perceive, reason and interact with 3D environments in real-time. Although reasoning and perceiving robots first come to mind, physical AI includes self-driving cars, adaptive assembly lines, robotic surgery, and smart buildings. Where traditional AI "thinks," Physical AI is capable of "thinking, then acting."

Rhoda AI raises $450 million at $1.7 billion valuation, unveils robot intelligence platform

Rhoda AI on Tuesday said it has raised $450 million in a Series A funding round that values the company at $1.7 billion and unveiled a robot intelligence system it says can handle the ⁠unpredictability ⁠of industrial environments. Rhoda AI on Tuesday said it has raised $450 million in a Series A funding round that values the company at $1.7 billion and unveiled a robot intelligence system it says can handle the ⁠unpredictability ⁠of industrial environments. Breakthroughs in artificial intelligence models that help robots understand language, interpret visual information and predict how the physical world behaves, combined with growing investment from major tech and robotics companies, are expected to drive robotics adoption. Newer compact high-performance processors are equipping robots for real-time perception and broad-skill operation. That momentum is fueling a race in humanoid robots specifically, ⁠led by Tesla, Figure AI, Unitree, Agility Robotics, and dozens of Chinese startups. However, industry experts caution that reliability, safety certification and ⁠cost will remain key hurdles for large-scale commercial deployment of general-purpose robots. Rhoda's new robot intelligence platform, FutureVision, works by first studying hundreds of millions of internet videos to learn how objects move and how the physical world behaves. It then uses that knowledge to constantly anticipate what is about to happen around it and translate those predictions into physical movements, a cycle it repeats dozens of times per second. The approach targets a longstanding problem in robotics: most machines perform well in controlled, predictable environments but struggle when ⁠something unexpected happens. The company expects to eventually license FutureVision to companies running robotic hardware and software platforms. Rhoda AI, which emerged from stealth on Tuesday, said its platform is designed to integrate with a wide range of robotic hardware, allowing manufacturers and logistics operators to deploy intelligent robots without rebuilding existing systems. Its funding round drew backing from Khosla Ventures, Temasek, Mayfield, Premji Invest and Capricorn Investment Group, among others.

Rhoda AI Raises $450 Million to Automate Manufacturing and Logistics | PYMNTS.com

By completing this form, you agree to receive marketing communications from PYMNTS and to the sharing of your information with our sponsor, if applicable, in accordance with our Privacy Policy and Terms and Conditions. The company is building a robot foundation model that combines pretraining on hundreds of millions of videos with closed-loop video predictive control to enable robots to operate autonomously in real-world environments, it said in a Tuesday (March 10) press release. Rhoda will use the new funding to accelerate the model's development and industrial deployment, according to the release. "By learning from internet-scale video and operating in closed loop, our systems are designed to adapt to real-world variability in ways conventional approaches struggle to achieve," Rhoda Co-founder and CEO Jagdeep Singh said in the release. "The goal is simple: robots that work in the real world, not just controlled lab settings." Rhoda's technology has already demonstrated autonomous operation in production environments, and the company's new funding will support additional industrial deployments and customer pilots, per the release. The company works with industrial partners across manufacturing and logistics, according to the release. Jens Wiese, managing partner at venture capital firm Leitmotif, which is one of Rhoda's backers, said in the release that in manufacturing, it has traditionally been difficult to automate tasks that have high variability. "What impressed us about Rhoda's approach is its ability to adapt to conditions that typically require human intervention," Wiese said. "Technologies like this can dramatically expand the scope of what can be automated, playing a pivotal role in reindustrializing mature economies." PYMNTS reported in November that physical AI is emerging as the next stage of robotics as advances in sensing, perception and large AI models give machines capabilities that traditional automation never supported. AMI, an AI startup founded by Meta's former chief scientist Yann Lecun, announced Tuesday that it raised $1.03 billion to help build AI systems that "understand the world, have persistent memory, can reason and plan, and are controllable and safe."