Runway has long been known for being a force in the creative industry with its AI tools.
The New York-based company built its reputation on video and image generation. Now, it is expanding into a new field: robotics.
Since 2018, Runway has focused on building AI models for artists and filmmakers. The company released Gen-4, its latest video model, in March, followed by Runway Aleph in July.
Both tools showed how far its world models had advanced. These models soon attracted interest outside entertainment.
Robotics and self-driving car companies began reaching out after seeing the potential of using Runway’s simulations for training.
“We think that this ability to simulate the world is broadly useful beyond entertainment,” said Anastasis Germanidis, Runway’s co-founder and CTO.
“It makes it much more scalable and cost-effective to train robotic policies that interact with the real world.”
Simulation
Training robots in the physical world is costly and slow. For self-driving cars, it can also be dangerous. One wrong turn on a real street can carry serious risks.
Runway’s technology is a safer and cheaper option as its models can recreate complex scenarios and test them repeatedly.
Engineers can adjust one variable at a time while keeping everything else constant. This level of control is difficult, if not impossible, to achieve in physical settings.
Germanidis explained it clearly. “If the car took this turn over this, or performed this action, what will be the outcome of that?
Creating those rollouts from the same context is really difficult in the physical world.”
Competitors
Nvidia, a leader in AI and graphics hardware, recently released an update to its Cosmos world models. These models also target robotics training.
However, Runway brings a different strength. Its background in video generation gives it unique expertise in creating realistic digital environments.
This sets it apart from hardware-focused companies.
Adaptation Strategy
Runway does not plan to create an entirely new set of models for robotics; instead, it will fine-tune its current systems. The company has also started building a robotics-focused team.
Investors support this approach. Runway has raised more than $500 million from firms including Nvidia, Google, and General Atlantic. Now, the company is valued at $3 billion.
Runway’s expansion is not just about entering new markets. It is about following a principle – simulation.
“The way we think of the company is really built on a principle,” Germanidis said.
“Once you have those really powerful models, then you can use them for a wide variety of different markets.”
Implications
The impact could be far-reaching. Safer self-driving cars could emerge after being trained in thousands of simulated crash scenarios.
Delivery robots may navigate crowded sidewalks more smoothly thanks to virtual training. Even hospitals could see robots that assist with surgeries or logistics trained in digital environments.
