October 5, 2010
When you have a 1,360 square mile military facility that needs patrolling, robots are the way to go, or at least that’s the position starting to be adopted by the US Army. Why? Simple: using mobile robots instead of permanent infrastructure (like fixed cameras and motion detectors) saves $6 million in up-front costs plus an additional $1 million a year in maintenance. The robots being used at the Hawthorne Army Depot (which stores tens of millions of cubic feet of low-level radioactive waste) are somewhat unsexily called MDARS (Mobile Detection Assessment Response Systems). They’re diesel powered, with a top speed of 20 mph, and they can stay on duty for 16 hours. Most of the time, MDARS do random fully autonomous patrols, using RFID tags to check the status of locks and gates. If they notice something out of the ordinary, a human can take over, using cameras, microphones, and speakers to interact with whatever needs interacting with. The bots are currently unarmed, so short of running someone over they can’t do much more than observe, although they are certainly capable of mounting remote controlled weapons. [ NNSA Press Release ] VIA [ Danger Room ]
October 5, 2010
This hair washing robot from Panasonic made the rounds last week, but I figured it was one of those things where good video was important, and DigInfo News came through today. One thing that the video doesn’t elaborate on is how the robot is actually scanning the user’s head shape in three dimensions to figure out just the right amount of pressure to apply, and it’ll use that shape to remember who you are and what shampoo and massage settings you like. In general, Panasonic is trying to create a robotic infrastructure to help Japan (and the rest of the world) deal with an aging population that’s going to need more and more support. As such, this system is primarily targeted at medical environments and not for installing in your bathroom. Yet. [ Press Release ] VIA [ DigInfo ]
September 30, 2010
iRobot has had a virtual monopoly on the consumer robot vacuum market since they introduced the Roomba in 2002. But with some new competition this year, there’s now a real perception problem when consumers compare a Roomba’s cleaning technique to that of Mint or the Neato XV-11, which are able to localize themselves, map a room, and clean in straight efficient lines. Irrespective of whether ’smart’ cleaning is more or less effective (and iRobot argues convincingly that its unstructured patterns do in fact clean better), Roomba’s pseudo-random behaviors seem less sophisticated and ‘dumber’ by comparison. The Roomba, which (I would argue) hasn’t seen a significant upgrade since the release of the third generation 5xx series in 2007, is going to need some kind of upgrade, because consumers are now expecting household robots to be smarter. This patent application, last updated in April of this year and unearthed by Robot Stock News, seems to suggest that iRobot is at least considering adding localization to their line of cleaning robots. The patent is for a “Celestial Navigation System for an Autonomous Robot,” and works by using (and stop me if you’ve heard this one) a projector to put IR spots on your ceiling that the robot uses to figure out where it is. Yep, sounds a lot (suspiciously a lot) like NorthStar, used by Mint. And just like NorthStar, if iRobot implements this system it’s going to mean that you’d need a projector in every room you want cleaned. The Celestial Navigation patent does talk about a bunch of potentially interesting features… For example, each room gets its own ID, so you could schedule your robot to clean specific rooms at specific times, and then monitor its progress on a remote. And there’s even the suggestion that the beacons will be powered […]
September 27, 2010
The kids at Ecole Polytechnique Federale de Lausanne (or EPFL) have been cooking up quite a bit lately, as this video demonstrates. Not only have they put together a scalable system that will let any flying robot perch in a tree or similar structure, but now they’ve gone and developed a platform for swarming air vehicles (with Linux, nonetheless). Said to be the largest network of its kind, the ten SMAVNET swarm members control their own altitude, airspeed, and turn rate based on input from the onboard gyroscope and pressure sensors. The goal is to develop low cost devices that can be deployed in disaster areas to creat ad hoc communications networks, although we can’t help but think this would make the best Christmas present ever. See for yourself after the break. Continue reading EPFL develops Linux-based swarming micro air vehicles EPFL develops Linux-based swarming micro air vehicles originally appeared on Engadget on Mon, 27 Sep 2010 12:14:00 EDT. Please see our terms for use of feeds. Permalink Make | source EPFL | Email this | Comments
September 27, 2010
When we first laid eyes on the Sarcos XOS military exoskeleton three years ago, its sheer power and dexterity left us in awe… but as you can see immediately above, that wasn’t enough for Raytheon. Today, the defense contractor’s unveiling the XOS 2, a lighter, stronger robotic suit that uses 50 percent less power for dropping and giving us several hundred pushups. Video and a press release after the break don’t specify the suit’s military duties (they’re focused on instilling the notion that the XOS 2 is a real-life Iron Man) but we can definitely imagine these causing some serious damage if Hammer Industries decided to weaponize that high-pressure hydraulic frame. Update: We previously stated that the suit didn’t need to be tethered to a power source for operation, but that information was incorrect. [Thanks, SmoothMarx] Gallery: Raytheon Sarcos XOS 2 exoskeleton – press pics Continue reading Raytheon revamps Sarcos exoskeleton, creates better, faster and stronger XOS 2 (video) Raytheon revamps Sarcos exoskeleton, creates better, faster and stronger XOS 2 (video) originally appeared on Engadget on Mon, 27 Sep 2010 09:01:00 EDT. Please see our terms for use of feeds. Permalink | | Email this | Comments
September 27, 2010
Have you ever felt uncomfortable asking people for money? So has Chris Eckert, so he built a robot to help him out, named Gimme: Gimme is a two axis numerically controlled sculpture that pans a room looking for people. Once found, the machine tracks a person, cajoles them into making a donation, and resumes scanning the room searching for potential donors. The sculpture is controlled by an Arduino Pro Mini. Stepper motors are driven by two Pololu A4983 Stepper Motor Driver Carriers. The microcontroller, stepper drivers, and sensors are all mounted on a custom circuit board made with Eagle CAD. While I absolutely love the construction of this robot, I’m not quite sure what to make of the bare eyeball… Part of me says “cute!” Part of me says “get it away from me!” But for the purposes of this robot, that’s probably perfect: first it draws you in, and then when it’s too late, you have to pay it to get it to leave you alone. Genius. [ Gimme ] VIA [ Make ]