Personal Appearances

I'll be making two important public appearances in the next few weeks. I'll be addressing the IVTT (Intelligent Vehicle Technology Transfer) workshop on Intellidrive - Vehicle Infrastructure Integration this week (July 30) in Washington DC. This is held at the Holiday Inn in Gaithersburg MD, next door to NIST. My presentation will be on "Network Centric Operations".

I'll also be making a presentation entitiled "Reducing Operator Workload" at the AUVSI (Association of Unmanned Vehicles Systems International) conference in Washington on August 10th. My company will have a robot exhibit at this conference with our UGV's, UAV's and USV's. (ground, air, and sea vehicles).
Read more!

Its MAGIC! Another robot challenge

OK, we liked the DARPA Grand Challenge, and its successor the Urban Challenge. And the X-Prize was very exciting, dramatic, and had a wonderful outcome. The Lunar Google X-Prize is a little daunting, and not for the timid, and will require LOTS of fundraising. What else is there to go after for inventors and robotics tinkerers? Now there is a new contest in town, and you have to like the name. Its MAGIC - Multi-Autonomous Ground robotic International Challenge. It's jointly sponsored by the US RDECOM (a research section of the Army) and DSTO, the Australian equivalent of DARPA. The objective is for a team of three robots to perform a task of mapping and searching an urban terrain. And they have to work together as a team. This sounds quite interesting, and I'll be keeping an eye on this. Sounds like a wonderful excuse to go to Australia.
Read more!

The Truth about EATR

It is always gratifying when something I'm involved in gets a wide reception around the Internet. In the case of EATR (see post, below) there has been a lot of speculation and just wild rumors about what we are trying to do with the Energetically Autonomous Tactical Robot. For example

Bot Junkie: Robot runs on Soylent Green
The Register: Robot to consume all life on earth as fuel
DARPA Funds Flesh Eating Robot

Which stories basically took the concept of a robot that finds its own food, and then wildly speculated on what that food might be, because in the press releases we just said "biomass" without specifying what that might include.

I'm one of the principles involved in EATR, as my company is one of the commercial investors in this project.

THE TRUTH ABOUT EATR

EATR was always conceived as identifying, selecting, and consuming vegetable matter (leaves, twigs, branches, grass) as its energy source. In particular, the easiest vegetable matter to grab from the robot's perspective, is what is already lying on the ground - dead leaves, dropped branches, etc. That material is gathered up and then ground into small bits that can be easily dropped into the burner. EATR IS A VEGETARIAN.

So EATR is far more like a robot cow or horse, grazing for its food to convert to energy.

And frankly, we don't have any means at all of converting protein into energy, so any sort of animal material would be useless.

As noted in the press release, there are four parts of EATR:
the computer system with the sensors,
a large robot arm,
an external combustion engine (in this case, the Cyclone power unit)
the mobility platform (i.e. a car or mobile vehicle like a golf cart)

In the first phase of EATR, all of the components are stationary and mounted on a platform. We will be "teaching" the system to recognize potential food, and be able to pick it up and place it in a container. That alone is a very difficult prospect. In this stationary state, we will be placing different materials near the robot and using various sensors to identify and classify the material, and then use 3D sensors to show the robot where the material is and how to pick it up. Right now we are using small tree branches about 1/2 inch in diameter.

The material will go into a chipper that will render it down into sawdust. What will work best is dry leaves, twigs, and small branches--in other words, dead vegetable matter.

The team chose a steam-type engine (actually an external combustion engine) since that was the simplest and most direct way to make energy out of found material. We can just burn the material and use the heat to make electricity, which is our aim.

So the main food source of EATR, what we will get the best energy from, will be dead plant material that has dried out. Animal biomass would not be usable at all.

We are also planning for EATR to have a helper. It carries one or more smaller "marsupial" robots in a garage in the back of the vehicle (see image). We are thinking that these smaller robots (like the Elbit Beagle Robot pictured) can assist the larger "mother" vehicle to gather material into a heap that would be easier to pick up. The small "joey" robots would be battery powered and recharged by the larger EATR main vehicle. This would help deal with small scale objects like dead leaves that may not be efficiently picked up by the large robot arm by itself.

For our next phase, if we get to do one. is to create a more sustainable fuel source. NASA refers to what we do as "In-Situ Resource Utilization" or ISRU. They are studying using processes to extract fuel, air, and water from local materials on the Moon, and Mars. It is obviously a tremendous advantage if you don't have to carry all of your supplies with you everywhere you go. Here on Earth, the EATR team is considering some sort of bio-mass fuel creation, either using alge to make some sort of oil or bio-diesel, or something to do with either methane or ammonia, both of which can be used in internal combustion engines. It probably would not hurt to suppliment the energy budget with some solar electricity.

What do you do with an EATR? We've come up with a list of ideas, which include long border patrols, pipeline inspection, long term climate studies, wildlife research, mobile observation posts, and forest management. Any project that needs long duration missions away from civilization would be a possible application.

Frankly, I'm as interested in other technological spinoffs as the possiblility of making fuel. Just solving the object recognition and manipulation tasks would have immediate benifits for tasks like humanitarian landmine removal, or creating a robot that could tend plants. EATR represents an serious advance in autonomous robotics, and has tangible benifits to the military and society in general. I'm proud to be a part of the project, and particularly delighted to be working with the calibre of talent that is involved.

Also, by the way, I created most of the artwork that you are seeing in the press on EATR, including all of the pictures on this blog. Dr. Jim Albus created the original robot design, and I created the drawings. The "Tree eating robot" cartoon was my idea.

Read more!

Latest Thoughts...

What I'm reading now

Old Man's War by John Scalzi. A new take on "Starship Troopers" where you can trade in your old body for a new model, provided you volunteer for military service in return. Some fascinating thoughts on the nature of consciousness and a truly different look at the future. Somewhat reminds me of the Reality Disfunction novels of Peter Hamilton, which are wildly inventive. Good book with several sequels.

DroidMaker: George Lucas and the Digital Revolution by Michael Rubin. Actually not about robots at all, but rather how George Lucas created entire new industries by accident in his quest to make digital filmmaking a reality. Discusses the birth of ILM (Industrial Light and Magic), LucasArts Games, and Pixar Films.

What I'm listening to now:

Tom Smith: The World's Fastest Filker. I have to blame Gerry Tyra, a buddy from the F-35 Joint Strike Fighter program, for getting hooked on Filk, which is a science-fiction fan's version of folk music. Tom Smith is funny, gritty, thoughtful, and witty, as well as having a fast turn of verse ("It may not make sense but at least it will rhyme", he says). His biggest hit, Rocket Ride, has become my personal anthem. Also good is "Superman Sex Life Boogie", "Rich Fantasy Lives", and "Creatures of the Night". He just seems to get better, and some of his latest stuff, like "Lars Needs Women", and "Dead Again" is really good. Trust me. You'll find something to like.

I apologize that this annoying "Read More" flag shows up when I don't want it to. I had to try and cut down some of the longer posts so you can browse the blog easily, but it even shows up in the short ones. I'll get some help on this. Sorry!
Read more!

Video of the Week - Self Driving BMW

My video of the week is from one of my favorite TV shows. I really only have time to watch one or two shows a week. Anyone who knows me can tell you that my favorite show now on is Mythbusters. That makes sense, the hosts, Jamie, Adam and Grant, are robot builders and fellow contributors to Robot Magazine. My "fav" second show is Top Gear, which is on BBC America here in the US. This is a very off-the-wall, very British show about cars. Fast cars.

In this video, Jeremy Clarkson meets a self-driving BMW 3-series. I've always found this video amusing because shortly before I saw it, I was at an intelligent vehicle conference, and had lunch with a couple of engineers from BMW, who flat out told me that BMW had no interest in autonomous vehicles, since they were "all about the driving".


Now in their defense, I understand that this car is from their testing division, and is particularly designed to drive the exact same lap over and over so that the BMW engineers can gather data -- its not intended to be any sort of technology for a product. The other amazing fact is that this car is driving blind. BMW is using a form of extremely high precision GPS mapping to keep the car on the track. It has no obstacle detection or avoidance capability at all. This is fine for controlled conditions and closed tracks, and admittedly gets the job done for testing. It would not do to mix this car in with other traffic. Anyway, it makes me laugh, and I get to talk about Top Gear. Did you see the time they put all these rockets on a Mini Cooper.....

I do have other shows that I like that are no longer on. I'm a huge fan of Thunderbirds, and of the japanese anime series Captain Harlock.


Read more!

THE CASE FOR SELF DRIVING CARS

I'm starting a new series here on self-driving cars, or if you prefer, autonomous automobiles (or robo-cars - what do you think they should be called?)

I've had many people tell me that they just were not interesting in a car that drives itself. What fun would that be?


My answer is quite a lot! From a selfish perspective, I live in Texas. This means I drive a lot on long highways, especially since my relatives are some distance away, and my children are in college out of town. I also spent two years commuting once a week 200 miles away, and a year commuting 120 miles a day (60 to and from work). That is a lot of time in a car. I would much rather be reading, sleeping, playing a game, or composing my blog that driving my car on these long trips. How would you like to push the 'autopilot' button on your dash and fold your seat back and let the car take you the next few hours on the superhighway?

That's my reason, but there are lots of others. The most pointed one is the simple fact that each year over 27,000 people are killed on US highways. Every year. A greater number of people are injured or maimed, resulting in long-term hospital care. Annual drain on the US economy is in the billions. Now imagine a car that will not crash. It won't fall asleep, won't run red lights, and can't drive drunk. What is that worth?

Speaking of which, the widespread use of autonomous automobiles could eliminate drunk driving as a hazard, a crime, or a problem. If your too soused to drive, the car takes you home. You don't even have to remember where home is.

Another point is that self-driving cars could eliminate driver's licenses, age limits, and handicaps. Are you blind? The car can drive. Too old? No longer a problem. Too young? the car drives itself, is a license required? As we are all going to get old and infirm some day, the allure of keeping your personal transportation, safely, is tempting.

Now how about saving the environment. Every day your car sits in a parking lot, doing nothing. I drive to work and park my car. My wife goes in the opposite direction and does the same. If we had a self driving car, it could take me to work, then return home and pick up my wife, and take her to work. Then it could return home and wait for us to get off. No parking lot issues at work, and the whole family only needs one car.

Rental cars would also take in a whole new meaning. In the anime "Ex-Driver", the animators explore a world of autonomous cars. You need a car, you call on your cell phone to a rental agency. Minutes later a car shows up (following the GPS in your phone) by itself and picks you up. You type in where you want to go and sit back. When the car gets to your destination, it waits for the next call in the area and drives itself to get someone else. Seeing a movie? Make a reservation for the right time and the car is there waiting for you when you get out. No parking.

So that's just a start. In later posts, we can talk about how this can come about, what the roadmap is to the fully self-driving car, and how changes to the infrastructure, cars, and drivers can be brought about smoothly and incrementally.

For more information, I'm a member of the IVTT (Intelligent Vehicle Technology Transfer) organization. See their website for more information.

Read more!

Terminator Anxiety? How to stop worring about Robots Taking Over the World

Do you worry about robots taking over the world? Are you concerned that the Terminator movies are prophetic and that the robot apocalypse is just around the corner.

Not going to happen.

Why? Because robots are increadably stupid and they don't have any mechanisms to become smarter. Basically, current artificial intelligence is not. Intelligent that is -- we are no closer to replicating the thought processes in even the smallest ant than we were in 1968 when Arthur C. Clarke conceived of HAL.


As part of my job I get to work with, see, play with, and go to conferences to talk about the smartest, most capable robots and AI systems in the world. I also have a lot of exeperience with supercomputers. I've built AI systems and spent years as an AI researcher. And I'm convinced that we are all wet when it comes to computers and robots behaving intelligently. We are no only not on the right road to building a self-aware computer, we have not even found the right concepts to discuss it.

You may have seen a lot of new articles like this one, that say that soon computers will be built that surpass the human brain in compute power.

That is complete and utter BS.

While it may be possible to equal a brain's worth of number crunching, we are nowhere near creating a system that equals the brain's total bandwidth and memory capacity - because we are using the wrong materials. Let's look at a computer. We store information in a computer in binary code - 1's and 0's, on or off states. It takes millions of those ones and zeros to just encode a picture, and that's just the visual information.

Our brains, on the other hand, are completely analog systems. A single channel, rather than being on or off, can convey a complete range of values with millions of possibilities. (NOTE: some neurlogical SME will jump in at this point and say, "Wait! Neurons are either on or off" -- that's the wrong way to say it. Neurons either transmit or don't (true) but when they transmit, they provide a range of data, not just a single value.) So rather than having a memory cell that has a single bit of data that is either on or off, we have a single value that has a whole range - and can be a color, a smell, a sound, a touch, a muscle memory, a texture, a weight, and so on. So the total BANDWIDTH of the neural system is enormous. So while the computer can process the data - you can't get that much data in or out as fast as a brain.

To avoid getting too long on this post, let me summarize in a simple statement.

Untill we grow our computers from analog components
and teach them rather than program them
and a robot can learn to identify and manipulate objects just like a baby
we are not doing Artificial Intelligence

and you have nothing to worry about.

Read more!

ADVICE FOR ROBOT CAREERS

I guess I need to create a FAQ section for this blog, as this is another of the questions I get a lot of when I'm out giving talks.

How do you prepare for a career in Robotics?


Like many fields, Robotics is multi-disiplinary. We cover a lot of ground. On my current robot development team I have a manager, a business development guy (with a banking background), a mechanical engineer, two software engineers, and a systems engineer. Later we will probably need a technical writer.

The real secret, especially at my level, are presentation skills. We spend a lot of time making presentations, to customers, to other departments, and to management. I probably do four to six a week. In my experience, the designers who do well are the ones with the best communications skills -- you must be able to talk in front of a group, to organize your thoughts, and to communicate what you want to do to others. If you can't do that, leading any sort of team is very difficult. While many robot designers are "one man bands", that really limits the type of projects you can do, and certainly you would have problems at any major company.

The next secret is to take advantage of contests. I owe a great deal to taking part in the DARPA Grand Challenge, a race of full-sized self-driving cars. There are a great many robot contests, from LEGO's to the full AUVSI contests for college teams. At whatever level, these are great learning experiences. I have been involved in BEST (Boosting Engineering, Science and Technology) and think they have a good, well rounded program that really provides a taste of being a real robot maker. I'm sure FIRST and the other high school programs are excellent, but I've only been involved in BEST.

For high school study, of course you must take all the math and english you can lay your hands on. As Sinbad says, "MATH IS POWER" and that is certainly the truth. Out here in the "real world" we generally look up all the formulas in the book and do all the tough math in computers. You will not generally encounter tough math problems unless you get into UAV design and start doing CFD (Computational Fluid Dynamics). I struggled all through math class but retained enough to be functional. We really use a lot of trigonometry, so learn those sines and cosines. If you also struggle in math, don't let this disuade you.

Two other courses that I have never regreted taking in high school were typing and drafting. I don't know if they still call it drafting -- its probably CAD or something like that, but I learned with a pencil, a triangle, and a T-Square -- yes, back in the Cave Man Days. These have to be very high up on my list - typing is critical for stuff like making this blog, and drafting, or mechanical drawing, is a tremendous skill in translating your thoughts onto paper and is the only way to talk to machinsts and fabricators.

The other skills in robots include lots of computer programming. I have a strong background in computer simulation, which I think is a good background for any engineer. Of course, simulations are just a type of computer game (we used to say that the only difference between a computer game and a simulation is who pays the bill). As a design tool, we use simulation all the time, and it is very valuable. I also rely alot on computer graphics, so if you find a good computer graphics or especially computer animation course, take it.

From there some sort of introduction to electronics would round out the education. We use a lot of microcontrollers, which are a different sort of animal from other types of computers. There are some very good self-teaching courses for these, and I'd suggest getting either the Parallax BOE kit or the OOBUG to learn how to deal with PICS (Programmable Integrated Circuits).

Well, I know that was a lot, but let me encourage those of you who are thinking about a career in robotics -- this is the way of the future and in the next few years this career field is going to explode in popularity.

Read more!

Life Sized Gundam invades Tokyo

Celebrating the 30th anniversary of the "reality mecha" anime video, toy, and video game franchise, Bandai built a full size, 60 foot tall Gundam Giant Robot in Tokyo.

Unfortunately it does not move, walk, fight, or launch itself into space.

But it is really cool.

For those of you who have not heard of it, Gundam is the long running series of anime (Japanese Animation) series about "realistic giant robots" that have spawned thousands of model kits, of which my son is an avid collector. There have been several different series, including my favorite all time series title, "War in the Pocket".



What a delight to see a full scale model take shape.

I've been sitting on an article about giant humanoid robots for some time now -- maybe its time to trot it out. I'll just say that I'm generally in favor of giant robots as a concept.

Thanks to Cinematical

UPDATE: there is a cool new way to see this robot that I have to say is very impressive from a lot of perspectives (pun intentional) see the Microsoft PhotoSynth version of this project. Let me know what you think!
Read more!

Code Snippet: Resize Images

I'm going to be posting little bits of code here and there that I think are useful. Recently, for this blog I wanted to resize a bunch of images from the 6MP that my camera puts out (3000x2000) into something more appropriate for web publishing, say 640x480.


My current programming language that I use is Python, which is useful for a whole bunch of stuff and is what I program all my robots in.

The script resizes all the images to be 480 wide and otherwise to keep the previous aspect ratio - we don't want to distort the images if they are in Portrait (taller than wide) format. If you want a differnt size just change the "480" to something else (say "600" to get all the pictures to be around 800x600). You can also perform a format change using this technique by changing the line
newPic = "sm-"+newPic +".jpg"
im2.save(newPic,"JPEG")

To be something else, for example PNG format

newPic = "sm-"+newPic +".png"
im2.save(newPic,"PNG")

This Python script is dependent on PIL (Python Imaging Library) and the Win32 libraries and is written using Python 2.5 but should work in later versions:

Python for Windows
PIL Python Imaging Library
WIN32 for Python Utilities

Here goes:

import Image
import os


ddd = os.listdir(".")
pics = []
for ent in ddd:
if ent.find("jpg")>0:
pics.append(ent)
print len(pics)

for pic in pics:
im = Image.open(pic)
print "OPEN: ",pic
imSize = im.size
print "Old Size:",imSize
imSize = (imSize[0]*480/imSize[1], 480)

im2 = im.resize(imSize,Image.ANTIALIAS)
print "New Size:",imSize
ns = len(pic)
ns -=4 #take off suffix
newPic = pic[:ns]
newPic = "sm-"+newPic +".jpg"
im2.save(newPic,"JPEG")
print "SAVED: ",newPic
print " "

print "FINISHED"
## end

NOTE The blogger software is removing all of the tabs and formats from this code. If you know python you know it works by indenting loops

You can find another person's code for this at this location

Resize Image Code



Read more!

My Magazine Addiction

I have a problem that seems to be totally out of control. I went out of town for just four days last week on a business trip. When I returned, I had 9 magazines sitting on my desk waiting to be read. That, of course, would be in addition to the four magazines I bought to read on the plane.

AAAAAGGGHH


I have to admit that I like magazines. Its not just that I write for one (Robot Magazine) but that I like reading them. It does help that several of them are free. Since I'm a chief engineer and make a lot of decisions on R&D, I get people sending me free stuff on desktop engineering, embedded computers, COTS products (Commercial Off the Shelf) and so on. On the personal interest side, I read Flying and Private Pilot and get AOPA (Aircraft Owners and Pilots Association). Popular Mechanics is just a guilty pleasure, although I wish their articles were longer. And I now get Aviation Week and Space Technology, a truly wonderful publication. To tell you the truth, I rarely read the articles in many magazines I get. But I do look at all of the advertising. As a robot designer, I need a lot of estoeric stuff for my projects -- lately I've been looking at hyperspectrial imaging systems and ground penetrating radar, for example. So I find a lot of my "stuff" advertised in the magazines I read. And a lot of it is only advertised in magazines. So I do get a lot of value and I hope this encourages all those people who send me free magazines to continue and to tell their advertisers that at least one person reads every single ad. Now I just need to summon the will power to not save. every. single. magazine. forever.

I think I'm going to need a bigger... trash can.

Read more!

Must Read Robot Books

No all robot books end with taking over mankind or enslaving the world. My particular favorites are the BOLO books by Keith Laumer, and since his death in 1993, then taken over by other authors like the incomparable David Weber and William H. Kieth. The book to start with is The Compleat Bolo, by the original author. Mr. Laumer has quite a sense of humor, and is the author of the Retief novels, which may be the funnest science fiction ever written.

What is a Bolo? The Bolos are self-aware, intelligent tanks, or really self-propelled land battleships -- because Bolos are big. Really Big - with dozens of tracks, multiple turrets, and incredible weapons like infinite repeaters, and the great fan favorite, the Hellbore Cannon. The ongoing theme of all of the Bolo stories is that just because you weight 200,000 tons and have the firepower of a planet does not mean that a self-aware tank cannot have honor, faith, or loyalty. Bolos are supposed to take the best aspects of knights in shining armor or "the Honor of the Regiment" and be mankind's greatest protector. The "typical" Bolo story has an ancient battle machine, long out of use, being resurrected to protect a planet or colony from invasion, usually at the sacrifice of the machine itself. Bolos represent a standard for robot that in many ways is far more realistic than Asimov's three laws, and illustrates how ethics have a place in the midst of warfare. The best of the later Bolo stories is the novel "Bolo Brigade" by William H. Keith, who really found the right voice to carry on from the late Mr. Laumer, and a truly interesting enemy in the Malach, a race of pack-hunting intelligent dinosaurs.

Read more!