Robohub

Peter Corke and his huge team at QUT are doing great stuff in this area. They have spot-spraying robots, that zap only weeds and fertilize only crops, and also a great device that just mechanically smashes up weeds one-by-one. Peter has an admirable ability to choose projects and deliver the goods on them. Robotics needs more like him. Interview with Corke on this subject: http://robohub.org/robots-zero-tillage-robotics/ Homepage with free stuff including books and MOOC: http://petercorke.com/wordpress/ edit: I'm not affiliated with QUT and have no conflict of interest. I'm an academic that does related work but not in agriculture.

That also reminds me of an interesting research paper last year about a "Machine Learning Approach to Visual Perception of Forest Trails for Mobile Robots". ( http://robohub.org/drones-recognise-and-follow-forest-trails-in-search… or directly http://ieeexplore.ieee.org/document/7358076/ and probably available on sci-hub )

So... your dishwasher, printer, washing machine, oven, and thermostat? All carry out "complex actions automatically" and are controlled by computers. No one would consider these "robots". With all due respect, there is no agreed-upon definition. (As someone with a PhD in robotics, CEO of a robotics company, & robotics journalist for 10+ years. Even my old academic advisor and I had contention around 3D printers.) If that's not enough, here's a proper source: http://robohub.org/robohub-roundtable-why-is-it-so-difficult-to-define… The best definition I've ever heard: A robot is a device that doesn't work (yet); as soon as it works, it is renamed (Vacuum robot => Roomba; clothes cleaning robot => washing machine; etc.)

> will consider an AI to be "generic" when it is able to apply training from one domain to an entirely different domain without any manual "mapping" from humans. For example, being able to decently play checkers after learning chess and being given a description of checker's rules. Want to butt in to this discussion, and post this link here for anybody who hasn't seen/heard about this amazing development from a while ago: http://robohub.org/artificial-general-intelligence-that-plays-atari-vi…

Hmmm...I must disagree: http://robohub.org/video-throwing-and-catching-an-inverted-pendulum-wi… It's fine if you don't find this amazing. I do, and I think many others do too.

The definition of pro varies! Here's a short rundown of 3D imaging: Laser triangulation systems (laser stripe + turntable) will give you extremely highly detailed models at close range. You could build one for under £100 and the accuracy can easily be sub-mm. Downside is it'll only work for a short range. Long range you're not going to beat a scanning LIDAR system like the sort that Leica offers. A comparable system from Faro (with a public price-tag) is around £30-50k depending on spec. For 'set and forget' scanning, these systems are great. They're also pretty much the only reliable solution for robust, kinda fast capture of large sites. Velodyne's automotive LIDAR (the one on the older Google cars) came on the market for around $75k. The future of 3D imaging is probably low cost MEMS scanning LIDAR, which is being bankrolled by autonomous driving. Short range you also have time of flight (Kinect 2) which is used in research and also by several large companies for commercial RnD (e.g. Ocado - http://robohub.org/wp-content/uploads/2017/02/SoMa-Ocado-RBO-Hand2.jpg - see the thing on the tripod?). My local second hand stores sells the Kinect 2 for about £25, but you need the USB adaptor (£30). Lots of companies make ToF sensors, but Microsoft really forced the price down. Then you've got Kinect 1-like systems (Primesense), which include Google Tango (and other clones from Asus, etc). I think the Intel Realsense does something similar. These are mono cameras and an IR…

these are all broadly a) tech b) news c) aggregator: https://bearblog.dev/discover/ https://paperlined.org/ https://lemmy.ml/c/technology https://engineeringblogs.xyz/ https://jimmyr.com/ https://techurls.com/ https://interestingengineering.com/ https://researchbuzz.me/ https://www.techmeme.com/ https://robohub.org/ https://www.bleepingcomputer.com/ https://www.dailyrotation.com/ https://raddle.me/f/Tech

And for those interested, another podcast with Nick Patterson, who is mentioned at the beginning of this one: https://robohub.org/talking-machines-ai-safety-and-the-legacy-of-bletc…

I think there's still a private race to the moon. Some companies, notably, Astrobotic and Moon Express still seem set on getting there and have business models even though they dropped out of the Google contest. http://robohub.org/glxp-update-penn-state-lunar-lions-now-in-3-spot/

This makes an Alexa device a robot, especially when integrated in a smart house. I don't like that. (It's an occasional source of disagreement I have with a friend.) I think in order to be a robot, it needs movement. Either it can move something that could reasonably be called "itself", or it has direct peripherals like an arm on the same "itself" that can move things around it. Thus giving an Alexa device a robot arm, or some robot wheels, and letting it move itself or manipulate things around itself, would make it a robot, even if most of the brains are in the cloud. Of course even with the requirement of mechanical motion, you could still argue a toaster is a robot. It's hard to define robot: http://robohub.org/robohub-roundtable-why-is-it-so-difficult-to-define…

Who is going to start publishing open source car firmware? I imagine that putting something together that would run some of the newer cars would actually be quite difficult. I'm guessing that writing enough software to just get a Tesla to turn on safely and pull out of the garage will be quite an undertaking. I can see a huge benefit from having FOSS versions of car firmware in the future. It could be especially important if you don't agree with your manufacturers choice of ethics for automatic driving (See the AI Tunnel Problem). http://robohub.org/an-ethical-dilemma-when-robot-cars-must-kill-who-sh…

Seems like people are fairly evenly split on this issue if you look at the poll results: http://robohub.org/autonomous-car-poll-will-you-miss-driving-your-non-…

What, nothing about NATO's JANUS? > JANUS performance has so far been evaluated by many collaborating partners at centre frequencies from 900 Hz - 60 kHz and over distances up to 28 kilometers in waters all over the world. > JANUS packet and bit error rates have been computed as functions of the signal to the noise ratio (SNR) and time spread over periods extending from hours to months. Signal correlation times have been computed and long-term experiments by CMRE(external link) in 2008 and 2009 have helped quantify robustness during variable environmental conditions. * http://www.januswiki.com/tiki-index.php?page=About+Janus Toolkits available under GPLv3. * https://www.popularmechanics.com/technology/infrastructure/a27247/nato… * https://spectrum.ieee.org/tech-talk/telecom/wireless/nato-develops-fir… * https://robohub.org/janus-creates-a-new-era-for-digital-underwater-com…

> people have been trying to control nonlinear systems since the 18th century without much luck. Not so. Nonlinear systems are not difficult to control as a rule, with appropriate understanding of servomechanisms. To see how extraordinary such control can be, look at this video: http://robohub.org/video-throwing-and-catching-an-inverted-pendulum-wi… > If society/economy/culture is a dynamical system that generates the power-law income distribution and this system cannot dissipate its "energy" and the exponent reaches some critical value then it will collapse or we'll have some kind of catastrophic phase transition. These things happen, but usually because the system is not fully understood, or there's a reason it isn't being controlled, not because the system is innately uncontrollable. Obviously this moves away from technology and into philosophy, because the question becomes -- not whether the system can be controlled -- but whether such control is consistent with democratic principles. > I don't think power-law distributions have 2 tails Of course they do -- for example, any power law rule involving a power that's an even number: http://i.imgur.com/HU3KZQ7.png In fact, strictly speaking, all power law distributions have two tails, just not in the same direction: http://i.imgur.com/tcPYrJW.png