Full Self-Driving, On A Budget

Self-driving is currently the Holy Grail in the automotive world, with a number of companies racing to build general-purpose autonomous vehicles that can get from point A to point B with no user input. While no one has brought one to market yet, at least one has promised this feature and had customers pay for it, but continually moved the goalposts for delivery due to how challenging this problem turns out to be. But it doesn’t need to be that hard or expensive to solve, at least in some situations.

The situation in question is driving on a single stretch of highway, and only focuses on steering, so it doesn’t handle the accelerator or brake pedal input. The highway is driven normally, using a webcam to take images of the route and an Arduino to capture data about the steering angle. The idea here is that with enough training the Arduino could eventually steer the car. But first some math needs to happen on the training data since the steering wheel is almost always not turning the car, so the Arduino knows that actual steering events aren’t just statistical anomalies. After the training, the system does a surprisingly good job at “driving” based on this data, and does it on a budget not much larger than laptop, microcontroller, and webcam.

Admittedly, this project was a proof-of-concept to investigate machine learning, neural networks, and other statistical algorithms used in these sorts of systems, and doesn’t actually drive any cars on any roadways. Even the creator says he wouldn’t trust it himself, but that he was pleasantly surprised by the results of such a simple system. It could also be expanded out to handle brake and accelerator pedals with separate neural networks as well. It’s not our first budget-friendly self-driving system, either. This one makes it happen with the enormous computing resources of a single Android smartphone.

Continue reading “Full Self-Driving, On A Budget”

A Drone For The Rest Of Us

As anyone who’s spent Christmas morning trying to shake a quadcopter out of a tree can attest, controlling these fast moving RC vehicles can be tricky and require a bit of practice to master. [Erik] wanted to simplify this a little bit so his children and friends could race with him, and the end result is a drone that only needs two inputs to fly.

The results of his experimentation with simplifying the controls resulted in a “speeder” type drone which attempts to keep a certain distance off of the ground on its own thanks to an extremely fast time-of-flight sensor. The pilot is then left to control the throttle and the steering only, meaning that [Erik] can use pistol-style RC controllers for these machines. They have some similarities to a quadcopter, but since they need to stay level in flight they also have a fifth propeller on the back, similar to an airboat. This allows for a totally separate thrust control than would normally be available on a quadcopter.

The resulting vehicle is immediately intuitive to fly, behaving more like an RC car than a quadcopter. This also required quite a bit of processing power to compute the proper roll and yaw from a single steering input, but after many prototypes the result is impressive, especially since it was also built to use FPV as a means of control. One of the videos below demonstrates this video, and looks extremely fun to fly, and we wouldn’t mind seeing a race with these types of speeders much like we saw in the past with a group of pod-racing quadrotors.

Continue reading “A Drone For The Rest Of Us”

Tracked RC Vehicle Is (Mostly) 3D Printed

While wheels might seem like a foundational technology, they do have one major flaw: they typically need maintained roads in order to work. Anyone who has experience driving a Jeep or truck off-road likely knows this first-hand. For those with extreme off-road needs the track is often employed. [Let’s Print] is working on perfecting his RC tracked vehicle to take advantage of these perks using little more than 3D printed parts and aluminum stock.

This vehicle doesn’t just include the 3D printed tracks, but an entire 3D printed gearbox and drivetrain to drive them. Each track is driven by its own DC motor coupled to a planetary gearbox to give each plenty of torque to operate in snow or mud. The gearbox is mated to a differential which currently shares a shaft, which means that steering is currently not possible. The original plan was to have each motor drive the tracks independently but a small mistake in the build meant that the shaft needed to be tied together. [Let’s Print] has several options to eventually include steering, including an articulating body or redesigning the drivetrain to be able to separate the shaft.

While this vehicle currently has no wheels in order to improve traction, [Let’s Print] does point out that a pair of wheels could complement this vehicle when he finished the back half of it since wheels have a major advantage over tracks when it comes to steering. A vehicle with both could have the advantages of both, so we’re interested to see where this build eventually goes.

Thanks to [Joonas] for the tip!

Continue reading “Tracked RC Vehicle Is (Mostly) 3D Printed”

Prototype Robot For Omniwheel Bicycle

For all its ability to advance modern society in basically every appreciable way, science still has yet to explain some seemingly basic concepts. One thing that still has a few holes in our understanding is the method by which a bicycle works. Surely, we know enough to build functional bicycles, but like gravity’s inclusion into the standard model we have yet to figure out a set of equations that govern all bicycles in the universe. To push our understanding of bicycles further, however, some are performing experiments like this self-balancing omniwheel bicycle robot.

Functional steering is important to get the bicycle going in the right direction, but it’s also critical for keeping the bike upright. This is where [James Bruton] is putting the omniwheel to the test. By placing it at the front of the bike, oriented perpendicularly to the direction of travel, he can both steer the bicycle robot and keep it balanced. This does take the computational efforts of an Arduino Mega paired with an inertial measurement unit but at the end [James] has a functional bicycle robot that he can use to experiment with the effects of different steering methods on bicycles.

While he doesn’t have a working omniwheel bicycle for a human yet, we at least hope that the build is an important step on the way to [James] or anyone else building a real bike with an omniwheel at the front. Hopefully this becomes a reality soon, but in the meantime we’ll have to be content with bicycles with normal wheels that can balance and drive themselves.

Continue reading “Prototype Robot For Omniwheel Bicycle”

Four Wheel Steering, Always The Option, Never The Defining Feature

A couple of weeks ago when it emerged that a new Tesla might have a four-wheel steering capability, our colleague Dan Maloney mused aloud as to how useful a four-wheel steering system might be, and indeed whether or not one might be necessary at all. This is hardly the first time four-wheel steering has appeared as the Next Big Thing on the roads. It’s time to take a look at the subject and ask whether it’s an idea with a future, or set to go the way of runflat tyres as one of those evergreen innovations that never quite catches on.

What’s your dream vehicle? If you’re like me, you have more than one. There in my lottery-winner’s garage, alongside the trail bikes and the mobile hackerspace, the dictator-size Mercedes and the Golf Mk1, will be a vehicle that by coincidence has four-wheel steering. The JCB Fastrac is a tractor that can travel across almost any terrain at full speed, and though I have no practical use for one and will never own one, I have lusted after one of these machines for over three decades. Their four-wheel steering system is definitely unusual, but that makes it the perfect vehicle with which to demonstrate four-wheel steering. Continue reading “Four Wheel Steering, Always The Option, Never The Defining Feature”

Fail Of The Week: Roboracer Meets Wall

There comes a moment when our project sees the light of day, publicly presented to people who are curious to see the results of all our hard work, only for it to fail in a spectacularly embarrassing way. This is the dreaded “Demo Curse” and it recently befell the SIT Acronis Autonomous team. Their Roborace car gained social media infamy as it was seen launching off the starting line and immediately into a wall. A team member explained what happened.

A few explanations had started circulating, but only in the vague terms of a “steering lock” without much technical detail until this emerged. Steering lock? You mean like The Club? Well, sort of. While there was no steering wheel immobilization steel bar on the car, a software equivalent did take hold within the car’s systems.  During initialization, while a human driver was at the controls, one of the modules sent out NaN (Not a Number) instead of a valid numeric value. This was never seen in testing, and it wreaked havoc at the worst possible time.

A module whose job was to ensure numbers stay within expected bounds said “not a number, not my problem!” That NaN value propagated through to the vehicle’s CAN data bus, which didn’t define the handling of NaN so it was arbitrarily translated into a very large number causing further problems. This cascade of events resulted in a steering control system locked to full right before the algorithm was given permission to start driving. It desperately tried to steer the car back on course, without effect, for the few short seconds until it met the wall.

While embarrassing and not the kind of publicity the Schaffhausen Institute of Technology or their sponsor Acronis was hoping for, the team dug through logs to understand what happened and taught their car to handle NaN properly. Driving a backup car, round two went very well and the team took second place. So they had a happy ending after all. Congratulations! We’re very happy this problem was found and fixed on a closed track and not on public roads.

[via Engadget]

Karting Hands-Free

Some of us have computer mice with more buttons than we have fingers, resolution tracking finer than a naked eye can discern, and forced-air vents. All these features presuppose one thing; the user has a functioning hand. [Federico Runco] knows that amyotrophic lateral sclerosis, ALS, or Lou Gehrig’s disease, will rob a person of their ability to use standard computer inputs, or the joystick on a motorized wheelchair. He is building EyesDrive for the 2020 Hackaday Prize, to restore that mobility to ALS patients. There are already some solutions, but this one focuses on a short bill of materials.

Existing systems are expensive and often track pupil location, which returns precise data, but EyesDrive only discerns, left, right, and resting. For these, we need three non-invasive electrodes, a custom circuit board with amplifiers, signal processing circuits, and a microcontroller. He includes a Bluetooth socket on the custom PCBs, which is the primary communication method. In the video below he steers a virtual kart around a knotty course to prove that his system is up to the task of an urban wheelchair.

EyesDrive by [Federico Runco] should not be confused with the HackadayPrize2015 winner, Eyedrivomatic, lead by two remarkable hackers, Steve Evans and Patrick Joyce.

Continue reading “Karting Hands-Free”