Machine Learning Robot Runs Arduino Uno

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”

Exploring Tropical Rainforest Stratification Using Space-Based LiDAR

GEDI is deployed on the the Japanese Experiment Module – Exposed Facility (JEM-EF). The highlighted box shows the location of GEDI on the JEM-EF.
GEDI is deployed on the the Japanese Experiment Module – Exposed Facility (JEM-EF). The highlighted box shows the location of GEDI on the JEM-EF.

Even though it may seem like we have already explored every single square centimeter of the Earth, there are still many areas that are practically unmapped. These areas include the bottom of the Earth’s oceans, but also the canopy of the planet’s rainforests. Rather having herds of explorers clamber around in the upper reaches of these forests to take measurements, researchers decided to use LiDAR to create a 3D map of these forests (press release).

The resulting GEDI (Global Ecosystem Dynamics Investigation) NASA project includes a triple-laser-based LiDAR system that was launched to the International Space Station in late 2018 by CRS-16 where it has fulfilled its two-year mission which began in March of 2019. Included in the parameters recorded this way are surface topography, canopy height metrics, canopy cover metrics and vertical structure metrics.

Originally, the LiDAR scanner was supposed to be decommissioned by stuffing it into the trunk of a Dragon craft before its deorbit, but after NASA found a way to scoot the scanner over to make way for a DOD payload, the project looks to resume scanning the Earth’s forests next year, where it can safely remain until the ISS is deorbited in 2031. Courtesy of the ISS’s continuous orbiting of the Earth, it’ll enable daily monitoring of its rainforests in particular, which gives us invaluable information about the ecosystems they harbor, as well as whether they’re thriving or not.

Hopefully after its hibernation period the orbital LiDAR scanner will be back in action, as the instrument is subjected to quite severe temperature changes in its storage location. Regardless, putting LiDAR scanners in orbit has to be one of those amazing ideas to help us keep track of such simple things as measuring the height of trees and density of foliage.

No Moving Parts LiDAR

Self-driving cars often use LiDAR — think of it as radar using light beams. One limitation of existing systems is they need some method of scanning the light source around, and that means moving parts. Researchers at the University of Washington have created a laser on a chip that uses acoustic waves to bend the laser, avoiding physically moving parts. The paper is behind a paywall, but the University has a summary poster, and you can also find an overview over on [Geekwire].

The resulting IC uses surface acoustic waves and can image objects more than 100 feet away. We would imagine this could be helpful for other applications like 3D scanning, too. The system weighs less than a conventional setup, too, so that would be valuable in drones and similar applications.

Continue reading “No Moving Parts LiDAR”

Citizen Science Finds Prehistoric Burial Mounds

What do you do when you have a lot of LiDAR data and not enough budget to slog through it? That’s the problem the Heritage Quest project was faced with — they had 600,000 LiDAR maps in the Netherlands and wanted to find burial mounds using the data. By harnessing 6,500 citizen scientists, they were able to analyze the data and locate over 1,000 prehistoric burial mounds, including many that were previously unknown, along with cart tracks, kilns, and other items of archaeological interest.

The project used Zooniverse, a site we’ve mentioned before, to help train volunteers to analyze data. The project had at least 15 volunteers examining each map. The sites date between 2,800 and 500 BC. Archaeologists spent the summer of 2021 verifying many of these digital finds. They took samples from 300 sites and determined that 80 of them were previously unknown. They estimate that the total number of sites found by the volunteers could be as high as 1,250.

This is a great example of how modern technology is changing many fields and the power of citizen science, both topics we always want to hear more about. We’ve seen NASA tapping citizen scientists, and we’ve even seen high school students building research buoys. So if you’ve ever wanted to participate in advancing the world’s scientific knowledge, there’s never been a better time to do it.

Bicopter Phone Case Might Be Hard To Pocket, But Delivers Autonomous Selfies

Remember that “PhoneDrone” scam from a while back? With two tiny motors and props that could barely lift a microdrone, it was pretty clearly a fake, but that doesn’t mean it wasn’t a pretty good idea. Good enough, in fact, that [Nick Rehm] came up with his own version of the flying phone case, which actually works pretty well.

In the debunking collaboration between [Mark Rober], [Peter Sripol], and the indispensable [Captain Disillusion], you’ll no doubt recall that after showing that the original video was just a CGI scam, they went on to build exactly what the video purported to do. But alas, the flying phone they came up with was manually controlled. While cool enough, [Nick Rehm], creator of dRehmFlight, can’t see such a thing without wanting to make it autonomous.

To that end, [Nick] came up with the DroneCase — a bicopter design that allows the phone to hang vertically. The two rotors are on a common axis and can swivel back and forth under control of two separate micro-servos; the combination of tilt rotors and differential thrust gives the craft full aerodynamic control. A modified version of dRehmFlight runs on a Teensy, while an IMU, a lidar module, and a PX4 optical flow sensor round out the sensor suite. The lidar and flow sensor both point down; the lidar is used to sense altitude, while the flow sensor, which is basically just the guts from an optical mouse, watches for translation in the X- and Y-axes.

After a substantial amount of tuning and tweaking, the DroneCase was ready for field tests. Check out the video below for the results. It’s actually quite stable, at least as long as the batteries last. It may not be as flexible as a legit drone, but then again it probably costs a lot less, and does the one thing it does quite well without any inputs from the user. Seems like a solid win to us.

Continue reading “Bicopter Phone Case Might Be Hard To Pocket, But Delivers Autonomous Selfies”

Spoofing LIDAR Could Blind Autonomous Vehicles To Obstacles

Humans manage to drive in an acceptable fashion using just two eyes and two ears to sense the world around them. Autonomous vehicles are kitted out with sensor packages altogether more complex. They typically rely on radar, lidar, ultrasonic sensors, or cameras all working in concert to detect the road conditions ahead.

While humans are pretty wily and difficult to fool, our robot driving friends are less robust. Some researchers are concerned that LiDAR sensors could be spoofed, hiding obstacles and tricking driverless cars into crashes, or worse.

Continue reading “Spoofing LIDAR Could Blind Autonomous Vehicles To Obstacles”

Knife Throwing Machine Gets The Spin Just Right

Despite how it might appear in bad action movies, throwing a knife and making it stick in a target is no easy feat. Taking inspiration from the aforementioned movies, [Quint] and his son built a magazine-fed knife throwing machine, capable of sticking a knife at any distance within its range.

Throwing a sharp piece of metal with a machine isn’t that hard, but timing the spin to hit the target point-first is a real challenge. To achieve this, [Quint] used a pair of high-performance servo motors to drive a pair of parallel timing belts. Mounting a carriage with a rotating knife-holder between the belts allows for a spinning throw by running one belt slightly faster. The carriage slides on a pair of copper rails, which also provide power to the knife holder via a couple of repurposed carbon motor brushes.

At first, the knife holder was an electromagnet, but it couldn’t reliably hold or release the stainless steel throwing knives. This was changed to a solenoid-driven mechanism that locks into slots machined into the knives. Knives are fed automatically from a spring-loaded magazine at the back as long as the trigger is held down, technically making it full-auto. To match the spin rate to the throwing distance, a LIDAR sensor is used to measure the distance, which also adjusts the angle of the aiming laser to compensate for the knife’s trajectory.

The development process was fraught with frustration, failure, and danger. Unreliable knife holders, exploding carriages, and faulty electronics that seemingly fired of their own accord were all challenges that needed to be overcome. However, the result is a machine that can both throw knives and nurture a kid’s passion for building and programming.

Continue reading “Knife Throwing Machine Gets The Spin Just Right”