Robotics

The previous two seasons were so far beyond any of the older seasons that I’m demanding you to watch seasons 6 and 7, before the world championship starts up again.

In the past, Battlebots was full of disappointing matches. The robots broke so quickly it was never really worth your time. Whenever they died, it was never obvious or interesting. Just one anti-climatic fight after another. That age ended with season 6, with 7 being even more impressive.

I think everyone needs to give combat robotics a second look.

I was aiming to use LLMs with robotics in an upcoming project, and needed to first verse myself in what is the current must-know techniques in the space. To that end I read a ton of papers and wrote this article to try and suss out the best parts of current state of the art.

I hope this helps people; I'd be thrilled to discuss much of this as well!

Stackexchange just became a Cloudflare site, which means it’s now an exclusive resource and also everyone’s data including usernames & passwords are exposed to Cloudflare. Please consider removing it from the sidebar or caution people about CF. Thanks.

A cool application of RLHF (Reinforcement Learning w/ Human Feedback - the same approach as what OpenAI used to train ChatGPT).

The authors trained an agent to fly FPV drones at a level surpassing world champions.

I have argued for awhile now that the probabilistic nature of LLMs can represent a form of context that, when applied, can be utilized for robotic applications.

Seems I'm not the only one that had this idea. While simple, Microsoft researched applied a high level control library to demonstrate LLMs (ChatGPT) developing robotic task code.

ROS maintainers discuss popular robotics control and navigation algorithms in use within ROS2. The associated discussion can be found here.

If you're looking for what to study or try applying in your own projects next, this is worth a look.

This is the routine thread where we discuss what you'll be working on this week! A cool robot? A computer vision project? Something cool in reinforcement learning? 3d printing a drive train? Let us know!

Maybe instead you're studying something, or reading a paper that just came out? Post about it!

It’s also okay to say “nothing” too - it’s great for your mental health to take a break!

Looking for help? Ask a question! See someone working on something cool? Ask them about it! No project is too small or too "newbish"!

In my first attempt at a long form technical post, I talk about a project where I had to use deep reinforcement learning to try and solve a robotics application. It worked ok - the post talks about my struggles, solutions, and what I'd change up in the future.

Link to the full paper

Hi! I'm a software guy and would like to start out doing some robotics. Before I go out and get a bunch of hardware. I'd like to practice the fundamentals.

I'm most comfortable with C++ and C# and dotnet and am pretty comfortable with game engines like Unity Unreal and Godot.

I've started out modeling a three-joint articulated robot arm that i can control through signals to the individual joints, like controlling a stepper motor.

My goal is to figure out a system where I can declare the shape of a robot like this (armature size, number of joints, offsets etc) to create a virtual model of the robot. I want to be able to send target coordinates and a basis rotation to that model and receive a series of signals back that will move the head of the robot to that 3d coordinate and rotation.

Now, I'm sure there are systems and packages that do all the math for this already, so what tools/libraries do you guys use to do modeling like this?

I want to see if I can simulate it in a game engine, and if that works out maybe ill try it on a toy :D

Thanks!

It uses its limb-like rotors and wheels to move in a wide range of ways, saving energy and making it more versatile.

You can read more about M4 in the research paper: https://www.nature.com/articles/s41467-023-39018-y

M4 was also covered in an article by TheVerge posted on /c/[email protected]

In all, the M4 can “achieve eight distinct types of motion,” and Caltech’s press release says it can decide what to do autonomously using artificial intelligence to survey surrounding conditions. Gharib says its abilities could be useful when shuttling injured people to hospitals or exploring other planets.

Stuff Made Here is an awesome channel with rather complex and fun robots and ideas.

This is a panel discussion with various robotics professionals from both industry and academia about how they got into their career, followed by open questions from the audience about navigating the robotics field and job market.