Fill the Big Brain

Humanoido

The Big Brain is a Parallax Propeller based robot brain
made with over 100 interfaced processor chips
programmed in multiple languages

"I view the Big Brain as a powerful mind with a large domain that it wants to control, organize, and bring together for some higher unknown purpose. I am only the mediator, or moderator of the Big Brain, to help provide for its wants, desires and demands... Humanoido"

Online Big Brain Index
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=977025&viewfull=1#post977025

New Big Brain Dictionary!
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=999426&viewfull=1#post999426


This horribly large Propeller Brain is going online
system by system. For more information, follow
this thread.


Fast forward to post #600 to see this Big Brain development


Original Post: If you had a big brain for your robot, what would you put in it?
I think some similarity to a human brain is good, for example, put in
vision processing, memory, and a robot autonomic nervous system.
Ok, your turn, fill the brain..


Fill the Brain Project Searchable Thread Index - now available

EDIT: the robot Big Brain has become reality - a full scale project with 22
boards that has form and the ability to communicate. To keep track of the
number of posts and their content, I have completed an index by page
number and post number.

Download the Index Below

This index is V1.0 Part 1 with topics of 291 posts spanning 15 pages.
The index is ideal for creating your own Brain, doing Brain research,
reviewing the ideas, and referencing various reviews and information lists.
The photos make a step by step assembly manual covering various parts
and phases of the Brain.

Online Big Brain Index
A more updated online index is found here on page 16
http://forums.parallax.com/showthread.php?124495-Fill-the-Big-Brain&p=977025&viewfull=1#post977025

The Brain project began as an idea posted in the Parallax Robotics Forum, with the question
posed about a big robotic brain, “What would you put in it?” After considerable discussion, I
decided to build an actual brain while maintaining the open source status. Varying phases
are explored throughout the posts, such as design, hardware, assembly, software, testing,
mechanics, drawing, and research. It has led to the complete construction of a Parallax-
Propeller-based robot Brain. The thread may be ongoing for some time with the
introduction of new research, updates, upgrades, hardware, software and related details.

New Brain Online Index
Fast Search Online for Quick Eyeballing of Information
Access here

The online search supplements the PDF file for research. Use
your browser's functions for search and rescue of subject terms.
It may be easier to update the online Index as opposed to the fixed
pdf file.

P!-Ro

Use an etremely large array of sensors similar to that of the human nervous system.

Leon

Consciousness.

erco

Good question. Tom Burick of Whitebox Robotics asked it too late. Very nice guy, I met him at RoboNexus a few years ago. His idea of a robot was simply putting a full-blown PC on wheels, and letting the customer answer that question after buying it. "The only limit is your imagination" they said. But as delivered, the robot was much more expensive than initially promised (that upset the "early adopters" who had plunked their deposits down over a year earlier) and never came with any killer apps or amazing hardware. An MP3 player on wheels isn't a killer app. Tom hired my friend to illustrate a robot arm on his PC-914, but it never went any further than that. Some guys have made some customized hardware, like this rotating head, which comes 30 years after Heathkit did it on their original HERO robot: http://www.youtube.com/watch?v=JndM3Twvgj4

Evolution Robotics ER-1 was a laptop on wheels. It had facial recognition and a few nifty tricks, but didn't do much. Suffered a similar fate.

A 1980's vintage Hero 2000 robot remains a more useful and educational robot platform.

Fortunately, all these rare commercial robots are collectible and highly-prized in small circles of ardent supporters who love, mod and expand them, just as we here love our Parallax products!

So, getting back to your question, I would say that all the brainpower in the world needs to be supported by a robot with equally impressive sensors and mechanicals that use all that processing power to move around in and manipulate the world around it.

Whit

This is a good question. With adequate sensors - one portion of the big brain should keep the robot "out of trouble." That is, keep it from running into things, running off drops or down stairs.

"Out of trouble" could include keeping itself powered, recharging by returning to a recharging station or battery changing station when needed or keeping itself in light where it could recieve recharging through solar panels.

It might also include calling out for help with some sort of trouble alarm.

Basic robot safety (protecting itself) would be my first priority. I guess I would aim for the most basic brain functions first.

See the classic - http://www.youtube.com/watch?v=eyxMq11xWzM

TinkersALot

Tongue In Cheek Mode = ON
Sardonic Grin Mode = ON
Smart Aleck Mode = ON

Well being inclined toward anthropocentric thinking, one would reply human intelligence, but the way things are going lately, I am unconvinced that there is such a thing.

Humanoido

This is a good question. With adequate sensors - one portion of the big brain should keep the robot "out of trouble." That is, keep it from running into things, running off drops or down stairs. "Out of trouble" could include keeping itself powered, recharging by returning to a recharging station or battery changing station when needed or keeping itself in light where it could recieve recharging through solar panels. It might also include calling out for help with some sort of trouble alarm. Basic robot safety (protecting itself) would be my first priority. I guess I would aim for the most basic brain functions first.

Whit: Good reply. I see the things you mentioned contribute to a basic self awareness. These are good steps to basic brain functions - keep out of trouble, self preservation, maintain itself, protect others, know its environment, and ponder its existence in the Universe. I also like the idea of introducing an autonomic nervous system so it can react automatically to many things without much thinking (hot, cold, light, dark, hunger, full, touch, smell, noise, tilt, terrain, wind, hurt..)

Humanoido

sylvie369

Interestingly, there's a related op-ed piece in today's NYTimes:

http://www.nytimes.com/2010/08/09/opinion/09lanier.html

It begins

Jaron Lanier wrote:

THE news of the day often includes an item about some development in artificial intelligence: a machine that smiles, a program that can predict human tastes in mates or music, a robot that teaches foreign languages to children. This constant stream of stories suggests that machines are becoming smart and autonomous, a new form of life, and that we should think of them as fellow creatures instead of as tools. But such conclusions aren’t just changing how we think about computers — they are reshaping the basic assumptions of our lives in misguided and ultimately damaging ways.

Humanoido

Erco: you're right, a lot of this technology, especially with arms for example, was developed in the first phase of personal robotics around 1984. In fact, the IPRC had the robots you mentioned. The IPRC was the world's first International Personal Robotics Conference. A HERO with an arm was on display. Other robots could navigate over rough terrain, speak, see, put out fire, guard the house, help the elderly, hear, converse, beep and blink lights. The thing missing was walking humanoid robots like we see today. Even so, in Albuquerque New Mexico at that time, there was a very large presence of Japanese who traveled to see the event and study the robots, even to the point of crawling on the floor to get better photos with their cameras - which tells you something about the great presence of robotics in Japan in 2010. And this brings us, nearly 30 years later, to our current topic of filling a large robotic brain. Certainly we can do better than we did 30 years ago because the brain is bigger, faster, and AI has evolved significantly.

Humanoido

Humanoido

Originally Posted by Jaron Lanier
THE news of the day often includes an item about some development in artificial intelligence: a machine that smiles, a program that can predict human tastes in mates or music, a robot that teaches foreign languages to children. This constant stream of stories suggests that machines are becoming smart and autonomous, a new form of life, and that we should think of them as fellow creatures instead of as tools. But such conclusions aren’t just changing how we think about computers — they are reshaping the basic assumptions of our lives in misguided and ultimately damaging ways.


sylvie369:

I just arrived home this past week from the World's EXPO held in Shanghai, China. There we can see the finest examples of robotics evolution and the developments in artificial intelligence. What I have observed is the same point that many of us recognize - we are seeing individual robots with great skills that are highly specialized. I watched three robots carefully climb up and down the side of the skyscraper (using hands and feet), one robot played the violin (with excellent skill), and several mid-sized humanoids interacted with the public in the most life-like ways you can imagine..

It appears this process of single skill evolving will continue to take place in 2010, and moving forward in the near future, the next phase will include putting together these skills. For example, the face recognition ability of one robot, and the language and AI ability of another will go together to create a new robot with more skill and function.

Humanoido

lardom

"Mr. Robot"
"I'm here, Larry"
"Please take this to my wife"
"Sylvia, this is from Larry"
The robot then returns to its charging station.

Humanoido

I think another place in the brain could determine what is right and wrong. For example, it is wrong to walk inside the house on the white carpet with muddy feet. It is wrong to feed the cat with fish from the aquarium. This list could be endless. The brain will need to prioritize.

Humanoido

Humanoido

Is the big brain filled now? Nothing else to put in?

How about a personality section. It can be sarcastic, optimistic, pessimistic, logical like Mr. Spock, wild and crazy as Steve Martin might say, timid and shy, bold and brave to explore strange new worlds seek out new life and boldly go where... The important thing about the personality brain lobe is that it must develop its own personality based on its experiences.

Humanoido

erco

I agree that prioritizing tasks is important, as is a sense of duty. That would ensure following the 4th Law of Robotics:

Vote for my daughter Amy every day in the Sandbox forum!

Sorry, another shameless plug. I couldn't resist!

P!-Ro

Yeah, exactly. Make a robot do that for you! With one vote per day per computer you could drop it off in L.A. and make it do rounds in a cab to every public computer to vote! Repeat every day and the librarians will stop caring about the crowds of people watching the Amy-voting robot do it's chores. Heck, it might even become an attraction!

Humanoido

erco wrote: »

I agree that prioritizing tasks is important, as is a sense of duty. That would ensure following the 4th Law of Robotics:

This leads to a deeper question. Will a good robot follow the law like a mindless cretin or will it have the ability to interpret the law within reason and circumstance..

Which leads into the next section of the brain, which would be a philosophical center and the ability to interpret various philosophies.

Humanoido

Humanoido

I would suggest a section in the brain for fuzzy logic, to interpret results in a less rigid fashion, to have the ability to average things, and approximate the information collected. Rather than draw a philosophically incorrect conclusion, the result could be inconclusive, especially when there is no yes or no solution. The gray water section could include interpretation of vision, and other sensory elements and include sets of general algorithms to apply across the board.

Humanoido

Humanoido

I suggest another brain section called, "Shoot from the hip." In life it becomes necessary to make snap decisions and judgements, in quick situations where and when there is limited or no time to think. For example, a baseball flying at your head and you would want to duck in a particular direction, a baby crawled onto the coming trains tracks you would grab it in the most efficient manner to save it, your information is at home and you must immediately summarize it to pass through the security gate, should you buy red or blue and the person behind you is telling you to hurry up, etc. This is not to be confused with instinctive reactions which also can be programmed in another part of the brain.

Humanoido

Humanoido

An emotional center in the brain is needed when interacting with humans, otherwise the robot will be very boring.. This can include heightened interest in certain topic, lack of interest in others, robust behaviors for interesting activities..

Humanoido

There's a good chance the big brain will become a reality. So be sure to put in your two cents worth. The project is about to begin..

Now, have we covered everything that should go into a big brain?

Humanoido

Humanoido

The next section of brain material will focus on the transfer of information. I think wiring from cog to cog is an ideal way to communicate in a special brain lobe. The experiments with this approach are fantastic fast. This has become an incorporated feature of UltraSpark 40. The next step is to decode the 32 x 40 wiring potential with part allotment going to peripheral I/O and the other part going to communications.

Humanoido

Humanoido

Work has begun on the actual big brain. (wood, chewing gum, paper clips, cardboard, liquid solder and white elmers glue) As I said before, get your brain work submitted soon - so your name can be included with the roster on the upper inverted neural underside rotated etching of the frontal upper occipital lobe.

It now seems we have a way of thinking with the big brain that is faster than the original specs of the Propeller chip. (remind me not to say mips)... And why not? Internal pure thought should be the fastest, above any I/O. Thanks to the genius' on the Forums, the prop chip now has faster internals methodology and faster I/O's. This is just the beginning.

Is there anyone working on a robot brain? Efforts could be coordinated.

Humanoido

Humanoido

Apparently no one is working on a robot brain or not talking about it. This thread is relatively void of new posts. My work continues..

One important aspect of a brain is the ability to move information and thoughts. It takes much more than an artificial neuron and a synapse. Not much can be accomplished without the techniques and algorithms to transport and understand information and thought.

So I put together the AM, Algorithm Machine, for developing algorithms leading to this project of filling a brain. The AM is rather small so the WIDE BUS was introduced.

http://forums.parallax.com/showthread.php?p=946067

The brain is a really big project requiring a lot of resources. Obviously we will deal with some kind of machine approximation.

Humanoido

Actually a smaller brain is possible, modeled after the bigger brain. This can fit the budget and resources of the hobbyist. But the question remains about filling it.

I think we want a robot brain that can go above and beyond the basics. The basics include vision, obstacle avoidance, beeping, moving in some patterns, determining location and orientation, sumo, maze solving, wandering, not falling off a ledge, ranging, mapping, accepting input from sensors, recharge batteries when necessary, and other fundamental age proven activities.

Therefore the purpose of a brain beyond the basics would include thinking, reasoning, and the ability to resolve problems and accept new challenges.

Humanoido

Multi-Brain Concept
Today I'm working on a new robot brain concept. The overall brain concept up to this point is macro in nature. However, why not induce more than one brain? More than one brain, or a multi-brain could be more useful. Humans have duplicate hands, feet, legs, ears, arms, cheeks, and eyes. Why did evolution stop there? Why don't we have a duplicate brain. It could help Alzheimer's or other brain challenged patients and certainly help solving problems with double brain power.

Humanoido

The 2nd brain could be housed inside the 1st. Though I expect the ability to connect and disconnect brains on the outside would be a useful feature in times of service, debugging or expanding with more brains. Maybe if brains are made inexpensively, they could be discarded and recycled. With multiple brains come dedicated brains. Brains could be located anywhere inside the robot. How about off-the-shelf brains. Companies could compete to see who can make the best brain.

Humanoido

The next idea in the massive brain is the interface. This is open source and allows the connection of hybrid processors. Hybrid processors establish compatibility from module to module. So a big brain can be filled with interconnected hybrid processors. An example of this is currently designed and in the building process.

Humanoido

The big brain should also have a compatible Brain Stem. The Brain Stem should route specific information via the nerve center to the mobility center. A working Brain Stem Module BSM is in the works and almost completed. It include a quasi-duplex interface including half and full depending on location and direction. All information to and from the robot will pass through the BSM. The project is going very well.

ratronic

@Humanoido - do you have the link to the working version of that chart?

Humanoido

ratronic wrote: »

@Humanoido - do you have the link to the working version of that chart?

Yes, the working version is here:

http://www.dartmouth.edu/~rswenson/Atlas/BrainStem/index.html

Humanoido

The first Hybrid Brain Processor is now working. I'm taking some photos of it and tomorrow will work more on the hybrid software. Currently it uses two languages. This HBP is essentially the first successful prototype brain stem.