Installation and hands-on records and adoption notes from the field. The numbers come from what we installed and measured ourselves.
To have a robot learn to walk on its own, you first need to install the simulators. This one article covers what Omniverse, Isaac Sim, Isaac Lab and MuJoCo each are, the official download addresses and the versions we used, the official specs and our PC's specs, the install steps and how to check them, and where we actually got stuck.
Read →
We attached Booster K1's reinforcement-learned gait to a museum guide simulation. Reading markers, walking up to each exhibit and giving a talk, three gaits trained from different seeds each ran the course and all reached the exit, with no falls. All of it is simulation, and whether the gait is good enough is judged separately, by the average and worst value over several trainings with different seeds.
Read →
Our humanoid education comes in three configurations: Start, Lab (standard) and Research (advanced). They differ in training days, workstation spec consulting, and the number of remote technical support sessions per year. Here are the differences and the questions to settle first.
Read →
K1 fits courses centered on walking, sensors and basic AI; T2 is for manipulation research with both arms. We put size and weight, degrees of freedom, onboard computing and battery side by side from the manufacturer's published specs, and set out how to choose by course goal.
Read →
Through V4, the arms only had a penalty for "stay still," and they drifted sideways while walking. V5 adds a reward for swinging the arms back and forth in time with the leg gait, and abrupt motion and torso wobble dropped slightly. We also record the result at 0.7 m/s and what we have not verified.
Read →
There is a method where you do not teach a robot "walk like this" step by step. It tries things and gets scored, and that alone lets it find how to walk on its own. We explain reinforcement learning with an easy analogy, leading into the result of applying it to our K1 robot.
Read →
Robot prices are published on the K1 and T2 product pages. Simulation practice seats, instructor and assistant training, technical support, and installation and operation support are quoted to fit your institution. Here are the items to sort out before you request a quote, in order.
Read →
Goods with an estimated price of KRW 20 million or less can be bought through a private (negotiated) contract with a single quotation, and above that the quotation method depends on the counterparty and the amount. The fiscal year for schools and national universities starts in March and the budget proposal is drawn up in January, so prepare quotations in the autumn and winter before. Based on the original text of the laws.
Read →
We moved a Booster K1 walking policy trained in Isaac Lab to MuJoCo. Walking forward and turning carried over, but sideways walking fell. Randomized training (v2) removed the fall but the body turned 195°, and v4, with a gait shape score, cut the heading drift to -11.6°. This gives the method and the results as they came out.
Read →
Student laptop specs and the simulator install come before the robot. In the 12-week example, weeks 1 to 10 run without a robot and the physical robot is used in weeks 11 and 12. Here are the preparation steps in five stages.
Read →
Installing Isaac Sim 5.1.0 and Isaac Lab 2.3.2 on Windows with the official commands stopped us in five places. This article gives the pkg_resources error, an access violation right after training starts, and the missing URDF extension 2.4.31, with the error text and fix commands as they appeared.
Read →
Yes. On one PC with a single GPU, Booster K1 walking trained in 32 minutes into a gait that does not fall, and v4, with the gait shape corrected, trained in 1 hour 40 minutes into a human-like gait. This covers the PC specs, the tools, the cross-check in another simulator, and the limits.
Read →
One GPU ran 4,096 K1 robots at once and produced a gait that does not fall in 32 minutes. But it was a shuffle of 15.2 foot steps per second. Retrained with a gait shape score, v4 walked at 3.6 steps per second. We also record where the install got stuck and the limits we have not yet checked.
Read →
The math depends on the number of laptops, not the number of robots. The simulators are free and run on the CPU alone.
Read →
You download a public walking AI, load it into a robot, and it does not walk. The file may not be what is wrong. This is what we learned by opening that file ourselves.
Read →
Surprisingly, reinforcement learning practice does not use it. When we opened the actual deployment code, it was clear when it first appears.
Read →
Gazebo, Isaac Sim, MuJoCo, Webots. You have probably heard all four names, but it is hard to find out how they differ. Here is how each one behaves.
Read →
One number inside a file may be the cause. It is called armature. We opened Booster's public files and checked.
Read →
No, you can. Isaac Sim converts it for you. Here is what the actual training code looks like.
Read →
The file is not incomplete. The URDF format itself has no field for them. We opened all three files for Booster K1, T1 and T2 to check.
Read →