A timeless pursuit of intelligence.
Chemical perception for machines.
Smell is where we begin.
Long before life could ask questions, it could sense what mattered.
A chemical trace could guide an organism toward nourishment or away from harm. Before brains, perception was already shaping how life moved through the world.
As we build new forms of intelligence, we return to that ancient relationship between sensing and acting. We believe machines should be able to encounter the world through its chemistry, and learn what those encounters mean.
Our work brings together sensing platforms, foundation models, and data infrastructure to connect chemical observations with context and experience.
Our initial focus is chemical sensing for autonomous systems. Over the next decade, we want this capability to become a bridge between machines and the living world.
In defense, we envision drones and ground robots that investigate hazardous environments before personnel enter them. Their observations could help locate chemical threats, assess contamination, and inform decisions that protect lives.
In hospitals, laboratories, and critical infrastructure, we envision a building immune system: an awareness of everyday chemical conditions that helps recognize hazardous changes and guide a response. For biosecurity, we aim to investigate whether chemical signatures can provide reliable early warning of biological hazards.
Across industry, a material’s condition could guide its repair, reuse, or recovery. The chemical profile of food could help reveal the effects of storage and transport. Environmental restoration could draw on observations that show how an ecosystem responds to intervention over time.
Our ambition also reaches into the sensory worlds of other species. We want to understand how animals and plants exchange chemical signals, and explore whether our tools can help us communicate in ways meaningful to their biology. Beneath the ocean’s surface, we aim to map chemical and sensory landscapes, revealing relationships between marine life and its surroundings.
Progress will depend on learning which differences matter. An unfamiliar trace may call for another measurement, a revised explanation, or an intervention. We want to connect instruments, robots, and laboratories so that systems can choose what to investigate next, test their interpretations, and produce knowledge that others can examine and build upon. This requires preserving the circumstances of an observation, making uncertainty visible, and letting new evidence change what was previously believed.
The longer we live with the world, the more deeply we should understand it.
Our founding team combines company-building experience with research experience at MIT and Institute of Science Tokyo.