Thread: Stuff in the news: Brain waves cut braking distance by 12' & Google's autonomous car

Bad luck if your brain panics at first...

The number of times I start to hit hard but realise - by the time my foot exerts pressure - that that is NOT the answer.

Same too the difference between an illusion or misinterpretation and the slightly delayed reality.

Not that I have ever had problems braking in time (despite my 1965 4-wheel drums without power-assist etc), any close situation has been due to other factors - namely me!

Originally Posted by brain wave stop article

This is uncomfortable, takes up to half an hour to fit, and the wearers have to wash the gel out of their hair afterwards. Smaller, more lightweight versions are in development.

The paper also mentions that wearers of EEG caps have to keep fairly still which is not always possible while driving, particularly when executing an emergency stop.

Dr Blankertz also said more work needs to be done on avoiding false alarms - to avoid the possibility that the machine could misread a drivers' brain signals and brake unnecessarily.

years ago, there was a brain wave device on newegg that promised that you could control your computer without needing a mouse.. the reviews were similar. some got it to work, but never better then mediocre, and if i remember right, false signals were also a huge issue. not to mention, there are a lot of things i think about while driving, i would hate to have any of those things acted out as driving maneuvers..

I'd expect the sensing issue to be resolved eventually. (And be remote.)


Sensual controls (like vision for the mouse) can be complex awkward because of the difference between raw sense signals and how the brain overrides them, though that is usually the eyes that override all else - like hearing, balance, etc.

Eye sensing is probably best done through conventional detection (as used for targeting by fighter pilots etc). (Not that I am up on their long-term eye damage and effects.)

Originally Posted by Fiberoptic

Can you build an autonomous car? Google did and crashed it. You can do better. http://cnet.co/oIJDcr

In my line of work, I find this sort of thing really funny. So quickly does the media jump to "Are they really all they're cracked up to be?" that we are instructed very specifically to under no circumstances talk to the media, and to be wary of idle conversation in public places. So easily the smallest things can happen, that push the robotics field back YEARS in the public eye from the media. If there was any injury, or heaven forbid a death involved, it would easily add decades to any sort of robotics on the road. I think people need to just chill out. No wonder google is being so secretive with things like this flying around. The public is way too panicy and anti-change to even begin to embrace technology on this level.

This kind of reporting is horrible, and cnet should be ashamed. This is almost as bad as the article about the UK cash cab car hitting a pedestrian while a producer was driving it off scene. The article made it sound like the US cash cab killed someone on-show. Forgot who wrote that one though.


Oh, and if you think you can do better, Darpa challenge is open, and there are many grants and different funding options available

Well said Malcolm!

As to reporting, I find that with almost all news. It's sensationalism, and usually intended to start a frenzy, though sometimes in cases like this, a preliminary report where people fail to appreciate the current state and development delays - like further testing, 5-year drug approvals (and of course their falsified results - viz: Tasers).

Again, in this case, as Soundman pasted, "Dr Blankertz also said more work needs to be done on avoiding false alarms - to avoid the possibility that the machine could misread a drivers' brain signals and brake unnecessarily.". That elimination is a very complex/long process. (It's like proving something. In principle (scientifically) you can't - you can only eliminate the the wrong stuff.) That issue was also what I was alluding to - basic ideas are usually easy, it's the practicality and safety that account for production delays. (Compare products on Intel's or GM's drawing boards. Electronics are often "generations ahead" of current releases.

Of course, yet again I am preaching to an audience that probably understands all to well. I look over the years and workhours of development on this forum - sometimes years for some "simple" (ha ha!) feature. (Yeah - that's easy, I can write that. How long? A few years... I said easy, not quick.)

Originally Posted by malcom2073

...

thanks for reminding me, it is so easy to forget how skewed the media can be. like the stock market issues at the moment. the news is preaching nothing but gloom and doom, and people are freaking out, all while the serious investors keep saying to stick with whatever you have--the market will recover.

lol at the darpa challenge.. most of us can't even setup a fe without problems

it is unfortunate that the google car was in a accident, but i am sure that it gave them some critical data that could help keep everyone safer in the long run.

LOL - stock market.
Alas we in Australia realised that is merely speculation and that life goes on, hence we avoided the last recession. (Actually it was our then Prime Minister that did not do as expected. Instead he gave 2 lots of ~$1,000 handouts to tax payers. Thence extra spending "despite alleged hard times" etc. It's strange that he was later ousted and replaced with our current Prime Minister - or is that just me LOL!?)
Of course, how do share prices actually effect a company's supply & demand & profit? If there is a substantial link, then economic lifestyle crashes are understandable. But it seems supply & demand is effected AFTER (big) investors remove their money just in case next year's forecast return drops from 15% to 13%. (What a pity shares are no longer the "investment" they once were - they have become a temporary overnight commodity; profiteering on speculation instead of company business. Not that ordinary investors can do that - they often have to meet minimum requirements, and can't sell shares just after some calamity or change their superannuation or share portfolio immediately or more than twice per year etc. Else they don't have the real-time access required. Remember the crash of '87 (yeah, I'm old) caused simply by pre-programmed dumps at -10%. How stupid - such a foreseeable event!)
But that is the world we chose to live in....


The Google car was unfortunate, but hence why all the testing. And input from as many experts as possible. (Experts meaning those with contributing knowledge (ie, Soundman!) - not those with qualifications or direct involvement. (Fresher views/eyes are often revolutionary! It's a thinking thing, not necessarily knowledge.)



As to problems with anything to do with IT....
My favorite question from stakeholders: "How long before you find the bug?". (Or "how long to fix" before you have even investigated the beast.)
And as they say, it's the last 5% that takes 95% of the time. If only the UNIX philosophy were universally applicable - "get it mostly functional" - but law suits etc forbid that.


Written and Authorised by My Opinion (which everybody has a right to know).


But enough. I've replied excessively.
I should instead get into things I intended years ago - like MOSFET buffered UIBIs, transient-free SMPS MOSFET switching, mods to MW728 battery protectors (else equivalent circuits or PICs), and some bluddy suitable single-chip dimmer suitable for LEDs and dome/puddle lights (or fans) etc.

I dunno why everyone thinks it is a conspiracy..
Somebody crashed it while in manual mode, far more plausible than anything else..
Besides if an automated car DIDN'T crash in testing that would be hugely concerning!!
We aren't that great as programmers as soundman pointed out if there aren't bugs there is something you have missed!

You wouldn't you Aussie recalcitrant. You didn't even know there was a Global Recession!
LOL! (Hi Stretch! See you soon?)

Programming bugs? By us? Ha - my old tacho & accelerometer powered on the first time (my own design and assembly/manufacture) and got thru all bug-free 37 menu levels. (Yeah - ok, that's when the bug appeared that prevented access to the other non-bug-free tens of menu levels. But hey man, I gotta brag!)
[Alas that was in assembler which I find easier to be bug-free than higher level languages.]
I didn't think much of it at the time, but in retrospect, I know that rarely does the hardware alone power up! And this used a dual-CPU 6502 (a unique off-grid second (dual) line 64-pin socket) with external 12-bit DACs (to drive a plotter or analog device) and RAM and UV-EPROM and... whatever, all on a "primitive" (misaligned) dual-sided PCB with manually drilled holes; and a keypad.


But there are 2 general IT rules:
1. It will have bugs. (No matter how much pre-checking.)
2. There will be the things or exceptions that were not thought of.
Either way, you will be fundamentally fertilised. (Though I must admit, I have proven very good at preventing both aspects. But eventually.... And the worst is probably the stupidly obvious yet trivial things that should have been included, but these days I get a laugh out of those (often in front of management LOL!).


BTW - You know the y2k bug was a conspiracy don't you? I've read it so many times (and as I stated above, news reports and opinions are so very accurate).
Not to mention global warming - predicted and tabulated over 100 years ago.....


Oops - I said I'd bug out. (Pun intended.)
Gotta do the UIBI FET thing - after my appointment at my unemployment agency. Maybe next year... (is that to be 2006?)

But good onya pointing out conspiracies. But if MythBusters is taken as proof, I'm not surprised. (I'm still waiting to see all te theory and practice revised after they allegedly disproved lighting being attracted to mobile phones LOL!)
So many "conspiracies" for things that are essentially known or solved - at least in certain circles.


Bugging out....