Three pillars of Contact Intelligence.
Contact Intelligence is the ability to manipulate the physical world through rich, sustained contact. Reaching it requires progress on hardware, data, and model architecture at the same time — advancing any one of them alone runs into the limits of the other two.
Contact-enabled hardware
Hands built to seek and sustain contact rather than avoid it — and to make what happens during that contact observable.
Robotic Platforms →Contact-rich datasets
Data that captures sustained physical interaction, not just the collision-free trajectories between contacts.
Datasets →Contact-native AI architecture
Models whose representations and objectives are built around contact, instead of treating it as noise to be suppressed.
Open Source →How the pillars connect
Hardware determines what contact can be produced and observed. Datasets determine what can be learned from it. Architecture determines whether contact is treated as signal or discarded as noise. Each pillar sets the ceiling for the other two, which is why we do not treat them as separate research programs.
Written work in progress appears in Notes and Publications.