Machine Learning Robot Runs Arduino Uno

September 14, 2023 by Lewin Day 7 Comments

When we think about machine learning, our minds often jump to datacenters full of sweating, overheating GPUs. However, lighter-weight hardware can also be used to these ends, as demonstrated by [Nikodem Bartnik] and his latest robot.

The robot is charged with autonomously navigating a simple racetrack delineated by cardboard barriers. The robot is based on a two-wheeled design with tank-style steering. Controlled by an Arduino Uno, the robot uses a Slamtec RPLIDAR sensor to help map out its surroundings. The microcontroller is also armed with a Bluetooth link and an SD card for storage.

The robot was first driven around the racetrack multiple times under manual control, all the while collecting LIDAR data. This data was combined with control inputs to help create a data set that could be used to train a machine learning model. Feature selection techniques were used to refine down the data points collected to those most relevant to completing the driving task. [Nikodem] explains how the model was created and then refined to drive the robot by itself in a variety of race track designs.

It’s a great primer on machine learning techniques applied to a small embedded platform.

Continue reading “Machine Learning Robot Runs Arduino Uno” →

Big 3D Printed Hand Uses Big Servos, Naturally

September 1, 2023 by Lewin Day 15 Comments

[Ivan Miranda] isn’t afraid to dream big, and hopes to soon build a 3D printed giant robot he can ride around on. As the first step towards that goal, he’s built a giant printed hand big enough to hold a basketball.

The hand has fingers with several jointed segments, inspired by those wooden hand models sold as home decor at IKEA. The fingers are controlled via a toothed belt system, with two beefy 11 kg servos responsible for flexing each individual finger joint. A third 25 kg servo flexes the finger as a whole. [Ivan] does a good job of hiding the mechanics and wiring inside the structure of the hand itself, making an attractive robot appendage.

As with many such projects, control is where things get actually difficult. It’s one thing to make a robot hand flex its fingers in and out, and another thing to make it move in a useful, coordinated fashion. Regardless, [Ivan] is able to have the hand grip various objects, in part due to the usefulness of the hand’s opposable thumb. Future plans involve adding positional feedback to improve the finesse of the control system.

Building a good robot hand is no mean feat, and it remains one of the challenges behind building capable humanoid robots. Video after the break.

Continue reading “Big 3D Printed Hand Uses Big Servos, Naturally” →

Next-Gen Autopilot Puts A Robot At The Controls

August 25, 2023 by Bryan Cockfield 19 Comments

While the concept of automotive “autopilots” are still in their infancy, pretty much any aircraft larger than an ultralight will have some mechanism to at least hold a fixed course and altitude. Typically the autopilot system is built into the airplane’s controls, but this new system replaces the pilot themselves in a manner reminiscent of the movie Airplane.

The robot pilot, known as PIBOT, uses both AI and robotics technology to fly the airplane without altering the aircraft. Unlike a normal autopilot system, this one can be fed the aircraft’s manuals in natural language, understand them, and use that information to fly the airplane. That includes operating any of the aircraft’s cockpit controls, not just the control column and pedal assembly. Supposedly, the autopilot can handle everything from takeoff to landing, and operate capably during heavy turbulence.

The Korea Advanced Institute of Science and Technology (KAIST) research team that built the machine hopes that it will pave the way for more advanced autopilot systems, and although this one has only been tested in simulators so far it shows enormous promise, and even has certain capabilities that go far beyond human pilots’ abilities including the ability to remember a much wider variety of charts. The team also hopes to eventually migrate the technology to the land, especially military vehicles, although we’ve seen how challenging that can be already.

Several video clips of a robot arm manipulating objects in a kitchen environment, demonstrating some of the 12 generalized skills

RoboAgent Gets Its MT-ACT Together

August 20, 2023 by Michael Shaub 6 Comments

Researchers at Carnegie Mellon University have shared a pre-print paper on generalized robot training within a small “practical data budget.” The team developed a system that breaks movement tasks into 12 “skills” (e.g., pick, place, slide, wipe) that can be combined to create new and complex trajectories within at least somewhat novel scenarios, called MT-ACT: Multi-Task Action Chunking Transformer. The authors write:

Trained merely on 7500 trajectories, we are demonstrating a universal RoboAgent that can exhibit a diverse set of 12 non-trivial manipulation skills (beyond picking/pushing, including articulated object manipulation and object re-orientation) across 38 tasks and can generalize them to 100s of diverse unseen scenarios (involving unseen objects, unseen tasks, and to completely unseen kitchens). RoboAgent can also evolve its capabilities with new experiences.

Continue reading “RoboAgent Gets Its MT-ACT Together” →

Angry Robot Face Is Less Than Friendly

August 17, 2023 by Lewin Day 10 Comments

Sometimes you just need to create a creepy robot head and give it an intimidating personality. [Jens] has done just that, and ably so, with his latest eerie creation.

The robot face is introduced to us with a soundtrack befitting Stranger Things, or maybe Luke Million. The build was inspired by The Doorman, a creepy art piece with animatronic eyes. [Jens’] build started with a 3D model of a 3D mask, with the eyes and mouth modified to have rectangular cutouts for LED displays. The displays are run by a Raspberry Pi Pico, which generates a variety of eye and mouth animations. It uses a camera for face tracking, so the robot’s evil eyes seem to follow the viewer as they move around. In good form, the face has a simple switch—from good to evil, happy to angry. Or, as [Jens] designates the modes: “Fren” and “Not Fren.”

[Jens] does a great job explaining the build, and his acting at the end of the video is absolutely worth a chuckle. Given Halloween is around the corner, why not build five to eight of these, and hide them in your roommate’s bedroom?

Video after the break.
Continue reading “Angry Robot Face Is Less Than Friendly” →

Hackaday Prize 2023: PAROL6 – A GPL Desktop Robotic Arm

August 15, 2023 by Dave Rowntree 8 Comments

Parol 6 is a 3D-printed six-axis robot arm created by [Petar Crnjak] as a combination of the principles from a few previous projects. Aside from a pneumatic gripper, each axis is driven by a stepper motor, with at least a few of these axes being driven through a metal planetary gearbox for extra precision and torque.

From what we can glean from the work-in-progress documentation, there are some belt drives on four of the relevant axes and a mix of NEMA17 format steppers driving either 20:1 or 10:1 reduction boxes. There appears to be a mix of inductive sensors and traditional microswitches used, but it’s not so easy to work out where these are placed. Continue reading “Hackaday Prize 2023: PAROL6 – A GPL Desktop Robotic Arm” →

When The Sojourner Mars Rover Nearly Ran LISP

August 7, 2023 by Maya Posch 12 Comments

During the late 1980s NASA’s Jet Propulsion Laboratory (JPL) was busy developing the first ever wheeled robot that would roam the surface of Mars. Due to the long round-trip times of any signals between Mars and Earth, development of the firmware that would control the rover was a major point, with the two teams occupied with the task each picking different levels of autonomy for the rover. In a retrospective, [Ron Garrett] who worked at JPL on the ‘more autonomy’ team describes his recollections.

Whereas [Ron]’s team focused on creating a rover that could be provided with high-level instructions which the sophisticated LISP-based firmware would use as guidelines to navigate and operate by, the other team pursued a more limited autonomy approach whereby a human driver would use explicitly plan out the route which the rover would follow before awaiting new instructions.

Perhaps unsurprisingly, the system requirements for running LISP and the additional uncertainties and complexities with the autonomous approach, as well as testing and validating the firmware, resulted in the Sojourner Mars rover featuring the latter approach, with straightforward C-based firmware. Most of Sojourner’s autonomy was limited to a home return function if communication with the lander was lost, which limited both its range and operations during its 85-day extended mission.

As [Ron] covers with examples from later missions, one advantage of LISP is that it allows you to send instructions which can be interpreted (e.g. to debug the system) without having to program in such functionality explicitly. With later Mars rover missions much more of this autonomy that [Ron]’s team pioneered was implemented, although C remained the language of choice for these later rovers.

Heading image: Ron Garrett standing in front of the Robbie prototype. Rocky III can be see in the lower left, and above him are Rajiv Desai and Robert Ivlev, two other members of the team. (Credit: Ron Garret)