Foundations
Cyberwave Foundations
Essays on the principles of Physical AI, robotics infrastructure, and the programmable physical world.
Durable arguments — not announcements.
Each essay isolates a condition embodied systems must satisfy.
- PrefaceMay 11, 20269 min read
Before the Loop
This series is an attempt to recover a systems view of robotics at a moment when the field is being flattened by demos, labels, and model-centric language.
Max LungarellaRead - Essay 01Coming Fri, May 22
Robotics Is Not What You Think It Is
The Foundation Series begins with a simple claim: robotics is not the assembly of advanced parts, but the engineering of coherent systems that sense, estimate, decide, and act in the physical world.
Max LungarellaNotify me - Essay 02Coming Fri, May 29
A Robot Is a System, Not a Pipeline
A robot is not governed by a pipeline of components, but by a closed loop linking world state, measurement, estimation, action, and world response.
Max LungarellaNotify me - Essay 03Coming Fri, Jun 5
Time Is the Hardest Constraint in Robotics
Latency, deadlines, synchronization, and jitter are not implementation details in robotics. They are part of the structure of correctness.
Max LungarellaNotify me - Essay 04Coming Fri, Jun 12
Why Robots Never See the World Directly
Sensors do not reveal the world as truth. They provide partial, noisy, time-bound measurements from which the robot must infer a state that is never fully given.
Max LungarellaNotify me - Essay 05Coming Fri, Jun 19
Estimating Reality from Fragments
State estimation is the hidden backbone of robotics: the recursive maintenance of a usable belief about the world from incomplete, noisy, delayed, and asynchronous evidence.
Max LungarellaNotify me - Essay 06Coming Fri, Jun 26
The Geometry of Intelligence
Robotic intelligence depends on structure: frames, transforms, constraints, topology, and the geometric representations that make action possible.
Max LungarellaNotify me - Essay 07Coming Fri, Jul 3
Where Software Meets Physics
Robots do not execute abstractions. They act through motors, compliance, bandwidth limits, contact, and the stubborn constraints of embodiment.
Max LungarellaNotify me - Essay 08Coming Fri, Jul 10
Stability: The Invisible Requirement
Modern robotics may speak the language of world models, diffusion policies, and vision-language-action systems. But none of that becomes operational unless the machine stays stable while the world pushes back.
Max LungarellaNotify me - Essay 09Coming Fri, Jul 17
Planning Is Not Enough
A plan can be elegant, optimal, and completely unusable. In robotics, decision-making matters only insofar as it survives uncertainty, delay, and execution.
Max LungarellaNotify me - Essay 10Coming Fri, Jul 24
The Limits of Learning in Embodied Systems
Learning can improve perception, representation, adaptation, and control. It cannot repeal latency, partial observability, actuation limits, or incoherent system design.
Max LungarellaNotify me - Essay 11Coming Fri, Jul 31
Why Most Robots Fail in the Real World
Robots rarely fail because one algorithm is weak. They fail because reality is variable, interfaces are unforgiving, and small mismatches across mechanics, sensors, software, power, and connectivity accumulate into system-level breakdowns.
Max LungarellaNotify me - Essay 12Coming Fri, Aug 7
When One Robot Becomes Many
The move from one robot to many is not a simple scaling problem. It is a shift from a single loop to interacting loops, shared state, coordination, infrastructure, and eventually networked embodied intelligence.
Max LungarellaNotify me - Thesis 13Coming Fri, Aug 14
The Rise of Physical AI
Physical AI will not emerge from models alone. It will emerge from systems that can bind reasoning, perception, control, timing, and embodiment into operational competence.
Max LungarellaNotify me - GlossaryComing Fri, Aug 21
A Glossary for the Foundation Series
A compact glossary of the core terms used across the Foundation Series, written to keep the series operational rather than merely conceptual.
Max LungarellaNotify me