Defining a task in Isaac Lab means specifying a Markov Decision Process - states, actions, transitions, rewards, and a discount factor. Seeing the MDP turns reward design from guesswork into deliberate craft.

A policy is the small neural network Isaac Lab trains - a function from observation to action, shipped as a checkpoint and then as portable TorchScript and ONNX files.

How NVIDIA's Omniverse, Isaac Sim, Isaac Lab, and Isaac ROS layer together into a single sim-to-real stack.

From Someone Who Has Stopped Trying

A word about one of the most likable cities in the world.

I've been filming thunderstorms from terraces and balconies for years. The same obsession that sends me out with a phone also ends up inside Unreal Engine.

On a May afternoon on the Costa del Sol, a pod of dolphins passed through the Strait of Gibraltar just off the La Alcaidesa golf course. This is what it looked like.

In 2019, before AR toolkits were mature, pointing a phone at your desk and watching something appear in the world felt genuinely strange. It still does.

A short screen recording does more for your colleagues than any written tutorial. Here's why it's worth making a habit of.

How our RYOT group brought augmented and virtual reality into retail — and genuinely changed how people shop.

A short observation on why the plants you want are hard and the plants you don't are not.

A pixel-art sun rendered entirely from a Houdini network — no painting, no hand-placed pixels.

A completely unnecessary game born from one seagull, one dog, and one very unfortunate afternoon.

A small team made one of the most visually striking games in years. It shows in every single frame.

My wife summoned me to the garden. What followed was an unexpected deep-dive into the surprisingly well-documented world of hedgehog droppings.

First post — a test of the blog setup and a quick hello.