Choose an edition
Edition: Business Professional. Applications, costs and business decisions.
Physical AI — the path to a beautiful tomorrow.
THE REPLICANT TRUTHA NEWSPAPER ABOUT PHYSICAL ARTIFICIAL INTELLIGENCE
Forecast · 2026-11-02 · CET
MORNING · 09:00
9.8°
Cloudy
Wind: N 2.9 m/s · 54%
MIDDAY · 12:00
12.4°
Clear
Wind: N 5.3 m/s · 11%
EVENING · 18:00
11.1°
Cloudy
Wind: N 4.1 m/s · 21%
Daily: 8.6–12.5 °C
Precipitation probability: 94%
All we can do is learn to be human while the rain falls.
Issue № 5 ·
rev. 1 · Archive
Robotics research, from training and touch to evaluating skills and practical limitations.
On October 29, Japan-based Forcesteed Robotics announced a feature for its developing FOS platform. A person describes a desired operation in ordinary language, and the local Luqua software agent selects an approach for each step: predefined rules, taught movements or a model linking vision, language and action. It then proposes a node-based work program for the company's LEIVOR controller.
Constructing a realistic virtual training environment can require extensive expert labor before a robot starts learning. Pranjal Aggarwal and Sean Welleck propose Gym-Anything, where coding agents build simulations and an adversarial audit checks whether their physics and tasks satisfy explicit criteria. The selection of physical work draws on US economic activity instead of focusing solely on familiar toy demonstrations.
Human fingers and robotic fingers have different geometry, so copying a hand trajectory does not guarantee the same contact with an object. Haozhe Zhang and collaborators propose TouchBridge, a combined hardware-and-software system with a wearable human demonstration interface and a corresponding robotic interface. Both capture visual and fingertip tactile information to establish correspondence across different hand designs.
An instruction such as clearing a table hides many steps: approach, grasp, move and release. Yet large robot datasets often store only the overall task instruction. Xueyuan Wei and collaborators propose RoboAtom-Data, which assigns consistent time-aligned labels to short, or atomic, actions so learning systems can recover the structure of longer behavior.
A conventional room map can show a robot where a door is without revealing which way it moves or how much force opening it requires. Dibyendu Ghosh proposes PhysVLMM, a map that stores probabilistic beliefs about articulated components. After a measured pull, the system updates its estimate, and a planner can ask whether an action is likely to succeed within a stated risk limit.
New research emphasizes structured data, measurements and testable results.
Concepts & terms
The key ideas in this issue, explained briefly.