Cheetah-Cub Is A Cat-Like Quadruped That’s The Fastest Bot Of Its Size

cheetah-cub robot

We’re still a ways away from electric sheep roaming the fields pretending to bleat but robotics researchers continue to look to nature for four-legged inspiration. Meet Cheetah-Cub, a European Commission-funded research project, out of Swiss University the École Polytechnique Federale De Lausanne‘s biorobotics lab, that’s about the size of a house cat.

As its name suggests, Cheetah-Cub takes its cues from feline morphology with strings replacing tendons and actuators sited in the legs to do the work of muscles. The result is a robot that runs like a cat and is, according to its inventor Alex Sproewitz, the fastest robot for its size (under 30kg). To look at it’s like a miniature and less scary version of Boston Dynamic’s terrifying Big Dog bot. The latter is likely faster, being much taller, but for a bot with a mere 0.15m leg-length Cheetah-Cub can really go some — hitting a max of 1.42m/s or almost seven body lengths per second.

The Cheetah-Cub researchers have been aiming for fast gait first, with the bot’s design, but do also plan to work on improving its rough terrain traversing capabilities — including Big Dog-style “stand-up capabilities” – as the work progresses, says Sproewitz. Building legged robots capable of dynamic locomotion in rough terrain is a big challenge on both “the mechatronic level, but also for control”, he adds. So as scary as these bots inevitably look as they scuttle about on their test walkabouts there’s no fear of us humans having to outrun any of them yet.

There’s also no danger of Cheetah-Cub heading for any kind of commercial implementation any time soon, of course. It’s pure research. The road to a future infested with mechanical animals requires a lot more robotics researchers to put their heads together in the interdisciplinary areas of biomechanics and computational neurocontrol.

On the question of the role biology plays when designing legged robots, Sproewitz said he distinguishes between bio-inspired robotics, which is what the Cheetah-Cub project is aiming for, and the more faithful copying of bio-mimicking robotics. Cheetah-Cub’s tri-segmented leg design is therefore a bio-inspired “blueprint”, rather than a direct mimicking of a cat. ”We tested a leg design with the proposed pantograph [three-segment] structure, and a second (even more successful) leg design where several additional features were merged into,” he says.

This same blueprint approach is how the researchers are approaching the bot’s locomotion controls. “Our implementation of a mathematical model of a central pattern generator (CPG) is a simplified version of what was identified in vertebrate and invertebrate animals. A full copy of e.g. a spinal cord would not be feasible: complex networks of neurons with very different functionality exist in the spinal cord of larger animals,” he says. “Many researchers dedicate their entire career in identifying fragments of those networks.

“Again, currently we apply relatively simple models of CPGs. We assume that CPG networks responsible for locomotion have evolved, but have been partially maintained from simpler vertebrates (like lampreys and salamanders) up to humans. Therefore: Cheetah-cub robot is a natural continuation of Biorob’s research with its Lamprey/Salamander robot, and the implemented CPG control.”

[Image: Biorobotics Laboratory, EPFL]

Ford using robots to stress test new vehicles

Vehicles being driven by robots is nothing new. We’ve seen that in Google’s self-driving cars, as well as prototypes from Lexus and Audi. However, Ford is putting the technology to a different use. They’re using autonomous robots to operate new vehicles for the sole purpose of stress testing them on Ford’s race course. One of

Read The Full Story

Sharp branching into robo-vacs, grow lights, e-whiteboards to offset slow LCD sales (video)

Sharp branching into robovacs, grow lights, ewhiteboards to offset slow panel sales

Remember the Cocorobo robotic vaccum that could send you pics of your newly-cleaned carpet? Sharp is going to be doing a lot more of that kind of thing soon, if a recent demo at its research center in Nara, Japan is any indication. Since the company has been losing gobs of money on its tepid LCD-panel business, it’ll soon be using some of that tech in completely different industries: for instance, converting powerful LED lighting from TV backlights to grow lamps, and touchscreen TV panels to interactive whiteboards. Sharp admitted to PC World that it needed to branch into other businesses since “rivals have been able to catch up from behind” to its LCD TV and mobile phone businesses — and judging by the drastic actions the company’s taken to stave off disaster lately, it’ll need to hustle those products to market, stat. Check the video after the jump to see some of the prototypes in action.

[Image credit: PC World]

Filed under: , ,

Comments

Source: PC World

BrickPi Is A Robotics Hacking Platform That Combines Raspberry Pi And LEGO Mindstorms

BrickPi

DIY micro-robotics is having a moment. The latest project to take the crowdfunding route — via Kickstarter — to build out a platform for playing around with robotics is called BrickPi. As its name suggests, BrickPi is a mash-up of the Raspberry Pi microcomputer, co-opted to act as the brains of the robot, plus LEGO Mindstorms sensors, bricks and motors for crafting its working parts. Firmware is written in Arduino, making it open and hackable. Indeed, the BrickPi makers have put their hardware designs and software source code online for download on Github.

The BrickPi extends the Raspberry Pi with a board that snaps in place over the Pi to connect it to the various LEGO sensors (such as touch sensors, colour sensors and gyroscope). This is then contained within a plastic case that is compatible with LEGO bricks so it can act as the base for building out the robot. An on board battery connector allows the robot to be untethered from a power socket so it can go roving.

The BrickPi is the brainchild of educational robotics company  Dexter Industries which also sells sensors for LEGO Mindstorms. The Kickstarter campaign has five days left to run and is approaching $90,000 in pledged backing from more than 1,270 backers — hugely above the original (modest) goal of $1,889.

Going the crowdfunding route sounds like it was primarily about building a community and getting the word out for Brick Pi’s makers but they have added a series of stretch funding goals to explain what they plan to do with the extra money raised. These include adding more sensors and ports to the device and creating additional libraries (in C/C++, as well as the original Python libraries) to expand programming options.

“We have a lot of plans for the extra funds raised and they all include improving the user experience and opening up the BrickPi to a wider audience,” says Dexter Industries’ John Cole. ” That mostly means putting together some sharp tutorials, and putting together more examples.  In my humble experience, where a lot of technical projects like this go wrong is when they have only 2 or 3 example projects.  Adults can think of a lot of projects and interesting ways to use the product, but kids have trouble with it, get bored, and move on.”

The original goal of the funding campaign was to bring the cost of the BrickPi down to $35 — making it the same price as the Raspberry Pi. BrickPi also pledged to write a library for the Scratch programming language, itself developed for helping kids to learn how to code. With projects like this, and mOwayduino — another robotics platform in the making due to go the crowdfunding route shortly — hardware hacking for creatively minded kids has never looked so easy.

Apple announces Anki Drive, an AI robotics app controlled through iOS

Apple announces Anki, an AI app for iOS

Apple is just starting its WWDC keynote this morning, but it’s already announcing something quite interesting: a new company called Anki and its inaugural iOS app called Anki Drive, which centers around artificial intelligence and robotics. The name, which is Japanese for “memorize,” features smart cars that are capable of driving themselves (although you can certainly take over at any time) and communicate with your iPhone using Bluetooth LE. These intelligent vehicles, when placed upon a printed race track, can sense the track up to 500 times a second. The iOS-exclusive game is available as a beta in the App Store today, which you’ll need to sign up for — the full release won’t be coming until this fall — and it’s billed as a “video game in the real world.” According to the developers, “the real fun is when you take control of these cars yourselves,” which we can definitely attest to — the WWDC demo cars had weapons, after all.

Follow our liveblog for all of the latest news from WWDC 2013.

Filed under: , , , ,

Comments

Source: Anki

iRobot and Cisco Ava 500 dispatches robot reps around your workplace

Despite everything Apple is expected to unveil at WWDC today, a corridor-mapping human scale telepresence robot probably won’t be on the agenda; for that, we’ll have to wait for iRobot and Cisco’s Ava 500. Built on the iRobot modular Ava platform, the video collaboration ‘bot mounts a Cisco EX60 telepresence screen – complete with HD

Read The Full Story

The RK-1 Is An Arduino-Based Mobile Robot You Control With Smartphone Swipes

climbing-rocks-at-the-Edinburgh-national-museum-960x450

London-based roboticist Evangelos Georgiou wants to offer an open-source platform for helping Arduino hobbyists take their projects mobile, thanks to a remote controlled robot called the RK-1 that combines a programmable Arduino microcontroller with apps for iPhone or Android, tank tracks, DC motors and Wi-Fi connectivity. It’s a project that could really help out with home hobbyists, or with education workshops and classes to get people young and old more interested in robotics.

The apps for the RK-1 will be free to download from the App Store and Google Play, according to Georgiou, and they use swipe touch gestures to manage changes in direction and forward/backward movement. It’s simple, and intuitive, but does look like it could be a very cool way for people to add an extra dimension to their product. And since Georgiou is following the example of other open source hardware hobbyist gadget sellers like Adafruit, there’s ample opportunity for cross-pollination with other similar projects, with built-in support in the ultimate companion app for sensors and breakout boards favored by the Arduino community.

Georgiou is a PhD student at King’s College London, and is also working full-time as a software developer at the school. His area of expertise is the impressive sounding “autonomous non-holonomic mobile robots implementing computer vision and advanced machine learning methods,” which basically translates to him really knowing what he’s talking about when it comes to building bots. His co-founder in the project Reetu Kansal is an experience assurance manager, and has been helping with project design and operation management as the RK-1 has developed.

Georgiou is seeking just £5,000 (7,800 U.S.), but already has stretch goals in place for £15,000 on up to £50,000, in a fit of optimism. Pre-orders of RK-1 kits start at £150 ($234 U.S.). It’s an ambitious product, but its founder has both the software and hardware know-how to make it happen, and this could be a very welcome component for robotics home hobbyists and educators.

MOwayduino Are Mini Robots Designed To Get Kids & Kidults Playing Around With Robotics

mowayduino

If you grew up in the 80s you might recall Big Track: a programmable kids robo-truck that could be made to perform a handful of actions, like moving a set number of units in a certain direction, firing a faux laser or tipping out the contents of its dumper truck at a pre-set destination (assuming your parents’ had shelled out for that accessory).

Well, Big Track mostly sucked but only because the concept was ahead of what the technology could affordably deliver. Playing with Big Truck meant repeatedly driving into walls as you underestimated the number of units required to negotiate the space between the kitchen table and the door. What the toy maker got right was that kids are easily excited about robotics. Fast forward a few decades and enter mOwayduino: programmable Arduino-based robot toys designed to be used in conjunction with mobile apps (e.g for radio controlling the device via the phone’s accelerometer) plus hardware add-ons — creating a rich environment for learning by playing around with hardware and software building blocks.

Or that’s the idea. At the moment, mOwayduino is at the concept/prototyping stage.  The Spanish company behind the project is apparently aiming to crowdfund the idea via Indiegogo. For now, you can register your interest via their websiteUpdate: mOwayduino’s makers say the Indiegogo campaign to fund production will launch in less than two weeks. “If we succeed, in three months, it will be on market. For people supporting the Indiegogo project, mOwayduino will be available at a special price,” the company tells TechCrunch. “If we exceed the money we need for the production, we will develop a graphical programming app for tablets.” It also intends to have an open API for users to develop their own apps.

The basic design of the mOwayduino robot looks a bit like a desk-mounted pencil-sharpener that’s wandered away from its duties. On the base there are three wheels to facilitate free, circular turning movements. Each robot then has various on-board lights and sensors (line sensor, light sensor, obstacle sensor, microphone, speaker etc) plus a USB-rechargeable battery that’s good for two hours of use.

Then, up top are expansion slots where additional modules can be plugged in to augment and extend its powers. For instance, there will be a vision system expansion module for capturing real-time images and sending them to a PC screen. (Just imagine the apps you could create to spy on your siblings!) Another planned expansion module adds Wi-Fi so the robot could be controlled from a mobile device or send emails when it has completed certain tasks.

The combination of a feature-rich basic robot unit plus the ability to augment and extend functionality — with support for programming via Ardunio IDE, Phyton, Java and the kid-friendly Scratch language, and the ability for multiple mOwayduinos to talk to each other and operate in sync — suggests this tech will easily kick Big Track’s ass. Or it will if it gets off the ground.

mOwayduino follows in the footsteps of other learning focused hardware such as the Raspberry Pi microcomputer. Key to the latter device’s success has been its low price tag so it will be interesting to see what price-tag mOwayduino will carry. Hopefully the base units won’t be prohibitively expensive so that kids can get their home-hacking on.

MindWalker brain-controlled exoskeleton puts the paralyzed on their feet

An eight week EC trial of a brain-controlled exoskeleton potentially promising newfound mobility to those with lower-limb paralysis will finish this week, with the project expected to spark a five year development path to a commercial version. The device, dubbed MindWalker, is the handiwork of a team at the Free University of Brussels, which has been working for the past three years on a motorized exoskeleton that can be controlled and navigated via brain impulses. Now, New Scientist reports, the European Commission will assess the results, having funded the project so far.

mindwalker_1

The MindWalker is made up of two companion projects, as the name implies. Arguably more straightforward is the exoskeleton side, which has been designed to support the weight of an adult, keep them balanced when walking, and adapt to different walking styles. An integrated brain in the walker itself helps spot obstacles that could present an issue for the user.

However, more complex is the mind-reading part, which the team refers to as the Brain/Neural Computer Interface (BNCI). A non-invasive system, using a dry EEG cap that doesn’t require messy gels or intrusive surgery, the technology cap is paired with a portable amplifier to make sure the computer gets the right signals. Previous methods have also included flickering diodes that gage intention to move by where the eye pays most attention.

mindwalker

The training process begins before the user has even strapped into the exoskeleton, however. The MindWalker team has developed a virtual reality training package which allows potential wearers to train their brains to get the most out of the BNCI link. As well as creating virtual obstacle courses to navigate through, the training system also includes a motion-actuated seat so that they become used to the sensation of being moved around by the motorized legs.

The dual development means that, even if mind control isn’t suitable for a particular user, that doesn’t mean the exoskeleton itself is out of reach. A more rudimentary control system – with pressure pads on the sides, triggered by rocking within the harness – to move each leg is also possible.

With brain control, though, there’s a lot more finesse up for grabs. The EEG system can apparently differentiate between the aim to move slowly or at speed, meaning pace could be controlled simply by thinking differently. Down the line, it could mean expanding the system to users with even less mobility, potentially including those who experience full body paralysis.

So far, the EC has pumped €2.75m ($3.6m) into the MindWalker project over the course of around three years. It’s still a long way out from commercialization, however, and the estimate is that it will take another five or so before a production version could be ready. By that point, project member Thomas Hoellinger suggests, the system could be a lot more aesthetically discrete, with less weight, smoother movements, and potentially even a frame that could be disguised under more traditional legwear.


MindWalker brain-controlled exoskeleton puts the paralyzed on their feet is written by Chris Davies & originally posted on SlashGear.
© 2005 – 2012, SlashGear. All right reserved.

Warrior Web from DARPA aims to boost muscles, reduce fatigue and injury (video)

Warrior Web from DARPA aims to augment soldier's muscles to reduce fatigue and injury video

The US military’s dabbled with full-on robotic suits in the past, but it’s now looking at a less convoluted, more energy-efficient approach. A project called Warrior Web from DARPA aims to enhance soldier carrying capacity and minimize injuries by distributing loads better, providing better joint support and “reapply(ing) energy to enhance motion.” Such a suit would be equipped with sensors to detect forces, and be able to fit beneath existing uniforms while consuming only 100W of juice. The US Army has nearly completed five months of prototype testing using a multi-camera motion capture system (see the video after the break) to develop critical tech. The next step will be to design and fabricate a suit ready for real-world testing, which should happen in the fall — assuming the program keeps its footing.

Filed under: ,

Comments

Source: DARPA