Over the past decade, AI has made incredible progress in the digital world — text, images, recommendations, and predictions. But in 2026, I believe the next real leap isn’t just smarter models. It’s intelligence that can sense, decide, and act in the physical world.
That’s why I’m building Physical AI.
From Software to Systems
My background spans embedded systems, mobile platforms, cloud infrastructure, and machine learning. I’ve worked close to hardware — sensors, networks, real-time constraints — and I’ve also built scalable cloud services used by millions.
One pattern keeps repeating: most AI systems stop at inference. They classify, predict, or generate — but someone else has to turn that output into action.
Physical AI closes that loop.
What Physical AI Means to Me
Physical AI is not just robotics. It’s intelligence embedded into real-world systems:
- Devices that observe the environment through sensors
- Models that reason under latency, power, and reliability constraints
- Systems that take actions — not suggestions
This includes edge devices, embedded Linux systems, MCUs, and hybrid cloud–edge architectures. The challenge isn’t only accuracy — it’s robustness, timing, and safety.
Why Now?
Three things make 2026 the right moment:
- Edge hardware is ready — NPUs and accelerators are now practical outside data centers.
- ML tooling has matured — deployment, monitoring, and optimization are no longer research-only problems.
- Real-world systems demand intelligence — energy, manufacturing, mobility, and infrastructure all need smarter control.
The gap between “AI demo” and “AI system” is where most interesting work lives. That gap is exactly what I want to explore.
What I’m Building
My Physical AI projects focus on:
- Sensor fusion and real-time data pipelines
- Edge inference with cloud coordination
- Decision-making systems that operate continuously
- Integration with Digital Twins for simulation and prediction
These projects are open, iterative, and built in public. Some will fail. That’s part of the point.
What This Blog Is For
This blog is a working notebook — not marketing. I’ll document design decisions, experiments, mistakes, and lessons learned while building Physical AI systems.
If you’re interested in where AI meets hardware, systems engineering, and the real world, you’re in the right place. Follow along as I build, break, and rebuild these systems in public.