Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
For the out-of-band Slashdot experience (mostly headlines), follow us on Twitter, or Facebook. ×

Comment: take from the aircraft/drone world (Score 1) 183 183

HUDs only make sense if you're truly a supervisor of the vehicle, not the control system.

In aircraft/drones, HUDs are OK cause the aircraft really flies itself, the pilot is there for emergency situations and what I call trimming (small adjustments).

In a car, you control everything--still want to end up in that ditch--pretty easy....sure go ahread....

Once we get real supervisory based cars, yes, HUDs make no sense other than wiz bang. And much like 8" touchscreens & phone integration in cars are whiz bang--are distractions currently.

Comment: ok, psychological effects, but (Score 1) 52 52

there's likely heath effects too. For example, those FPV drone goggles, with the dual 2.4 antennas and receiver, hitting a 500mW transmitter are likely not that good being 1/4" from your head, pressed on your temples, with a 500mAh battery pressed against the back of your head too.

Comment: Re:Flexible Automation is Hard (Score 1) 45 45

Viola, Foxconn just realized there's a huge capital investment for robots.

Also, they realized that the current industrial robots can't do much--the vendors promised too much and require maintenance by a MS or PhD. Where as maintenance for a human is called dinner (simple).

Robots need encapsulated, standardized designs to be useful. The automotive industry sort of proves it.

Comment: Re:Japanese Paradox (Score 1) 38 38

The scenario you're talking about is a short term scenario. Yes, robots are taking over the unskilled labor jobs... mainly because unskilled labor cannot keep up with the manufacturing demand. Look at Foxconn, they can't keep up with the manufacturing of iPhones. They have to automate, much like P&G did for toothpaste and quaker oats. The automation will make the quality issue moot and keep costs & demand manageable.

As robotic automation gets more refined from this (and it will), that's the time we need to find how to get the unskilled labor workforce... trained for higher skilled jobs. And the bar is essentially raised. It's all about buying time.

Moving unskilled labor into skilled labor jobs and such doesn't happen overnight. Hence, the real problem is those that fall through the cracks in the short term (some go back to school and make out OK)--so, how do we either move them out of obsolescence faster or re-engineer their job faster to maintain their work value? In the old days of unions it was either yearly OTJ (which companies don't provide anymore) or strong arm protection (to the point of being freeloaders) aside from basic education. None of that 'really' exists today as it's a entrepreneurial free for fall. Maybe online education/MOGs will be a key, but the clock is ticking and people are getting left behind.

Comment: Re:maybe robots can fly the drones (Score 1) 298 298

Maybe it's because NYT is a newspaper, which sells ads...
And there's a movie about a stressed out drone pilot that released last month. There's always a tie-in to some sell....

====

Really, drone pilots are just that, extensions of a drone, hence are forced to act like drones during missions and need to basically fly the 'entire plane'. And that's stress by boredom--just look at airline pilots... same routine, same times, same crew, same daily grind, and usually ends with something that has to do with drugs & alcohol.

Sure today's autopilots can really carry out an entire mission and allow a pilot to really be a supervisor (and not a pilot), but that creates a whole new set of problems. Problems that the military can't, nor doesn't want to handle. Most aircraft are piloted & manned, hence our entire infrastructure, regulations, training, O&M and essentially development approach is currently pilot centric, hence R/C of UAS is the closest DoD will every get to 'real' drones.

Autonomous UAS or even full-supervisory UAS (no pilot training, no real 'pilot') is just scratching the surface. No organization knows how to handle that aspect at scale-- I can surely tell you (since I'm one of the guys building such as system).

Comment: Re:"Crunch Time" == Bad Project Management (Score 1) 336 336

Could it also be improper scheduling and management too? Every game company I've met has a dedicated s/w dev manager(s), follow some well known process, hire skilled/talented folks, good industry partnerships, yet, they are always overbudget or over schedule. That says a lot.

Comment: Re:Audio (Score 2) 227 227

Detection of a drone is much like viewing the sky. It's nearly impossible to spot something.

Unless there's a RF tag, standard audio config (common prop pitch), visual tag, or something that the vendors add to their drones and is a well known... i.e. a standard, the only way to detect drones are by intelligent multisensor systems. Humans are a great example... and that means time consuming, very expensive and very complex. And they'll (like humans) still have false positives on order of 20% or more. Currently anyone can change the above parameters, there is no standard. It's no different from an off-road car--you can basically go anywhere. Where as a typical car, you can't (must use highway system) and it's illegal to modify w/parts deemed off-road. A standard and framework is in place and the traffic cops focus on the 2% of folks that put illegal parts on their cars to go off-road...

As a drone researcher and now looking into this detection problem, now [illegally] popular over some of the most visited entertainment spots around the world, there's 2 aspects to consider: detection and countermeasures. Detection is somewhat well known, there's active research in the military field. But again false positives are a huge problem and why manned flight uses deterrents instead (i.e. instead of detecting birds, use audio deterrents & scare them away). Countermeasures is a whole other problem: so far all solutions either remove power (falling drone--not good), take over control--aka YOU assume legal responsibility (not good), or trap it with a net which again you are responsible to getting it to the ground safely (not good). The vendors can add some deterrent mechanism, much like the mobile phone industry can locate you via police warrant, but that means we need to ensure the firm/software does what it says--a certification or standard is needed. Likely another reason they haven't--they would too assume legal responsibility if the failsafe... failed. Vendors are currently all about fun features and applications. Safety/security has taken a back seat unless you're homeland security. And a lot of these solutions are not well tested, and really addressing edge cases or closed environments. The military solutions don't work well--they typically don't consider collateral damage.

The drone industry is pretty much in the early Internet days, pre-netscape, more like usenet. Standards need to be created and that will block out 98% of the current drone violators. Then you can let the security guys focus on the 2% "professionals" and everyone will be happy vs paranoid.

It's really no different from manned flight issues (someone taking over a plane)... except it much harder as the UAVs are smaller, more agile, and in my cases-- fully autonomous. Considering we can do two-button-touch "launch, run, land"... on multiple UAVs at the same time--yes, we are entering a brave new world folks.

Comment: Re:There can be only one. (Score 1) 443 443

If emacs is an IDE, then Linux is a robot.

I choose the tools that are great at one thing and not at all things. I want tools that are not jack of all trades, master of none... as the saying goes. That because nowadays standard communication interfaces and formats are everywhere (we've reached parity).

Much of the excitement we get out of our work is that we don't really know what we are doing. -- E. Dijkstra

Working...