Generation II frame / on the droid trail.
My thoughts on droid robots:
I am of the belief that the perfect droid is an advanced robot that has a function or functions that are useful. Using the computer industry as an example think of the microprocessor chip. When it was originally invented the personal computer was not the application that was thought of. It was intended to replace complex logic in automation applications like cash registers and vending machines. Their was a need for tools to develop these applications so the next logical step was to build a simple computer to write and test code and to program other small computer / controllers. This is the chicken and the egg philosophy. Soon it was discovered that these simple computer had some use as a souped up calculator, word processor and even in gamming. Some visionaries like Dan Brickland starting developing very specific and very advanced applications like his Visi Calc (Visible Calculator) that proved that personal computers could do something that was extremely useful. This opened an entirely new market of the office machines. Later other areas developed, like in the areas of CAD, software design, communications and yes gaming. This applications were referred to as killer apps. The defined the computer as a unique machine that can do things that no other appliance can do.
Everything that is built and sold must have, or should I say should have a reason for its existence. The personal robot, PR, industry is no exception to this rule. Today the PR industry is growing very fast, however, the are no killer apps for robots. Once again I am talking about non-industrial robots. As an exercise look on the internet for robots and see what is available. Look for what sort of applications are available. Ask what is that this device can provide you that is of any importance at all? You will soon discover that there is very little useful functions at all. Even device like robot vacuumed cleaners create very much interest. So the question comes down to this why should you as a smart consumer invest your limited amount of resources in time and money in robotics? I will begin by challenging each and every one that reads this to a simple test.
First ask yourself, are you an inventor? An inventor is generally a person that is thinking all of the time about how to create something new or how to use something more common to do something that it was not intended to do. There have been instances of individuals that invented only one thing in their lives, but this is not what I am talking about. The real inventors know who they are. The like to take notes and sketch solutions to to various problems. They see many obstacles as a challenge that needs a solution rather than a reason to quit. Ask you self this have you ever done something that others told you could not be done. Many time others resent this and if you want to not ruffle any feathers you might not do this again. These people are very toxic to you and can place a great drain of your physical and mental resources. I meet a lot of very smart people as I got to various places. I can see great intelligence in some of these people. I get really sick when a person that can create a computer routing program is working in a fast food restaurant. Their parents, teachers and friends have created their place for them and are expected to live there. These people could be the inventor of the next technologies and have been trained to fail. I witnessed multigenerational cultural degeneration. This is where each generation is less skilled or socially developed then the pervious generation. The child can not find work doing anything while the grand parents were college graduates.
Another thing that you can do is to keep journals. This is not a personal book with what you did with who last night. This is a creativity book with things that you have though of and may never act on. I have been keeping such books since junior high school. I look at some of these and laugh about thing that I had though of that did not exist then and are common place today. As a class project I was asked my my drafting instructor in 8th grade to thing of a consumer product that was not invented and to do several drawings of it. I though of a device that had a color television (portable) and used plastic disk that had movies recorded on them. The scanning device was a helium neon laser (did not think lasers would be solid state). Each disk was in a sealed plastic case and hung on a key ring. When a move was to be viewed the case was slipped into a slot on the television. Since the disk case was rectangular the key ring and other moves could be left hanging. My disk was two inches in diameter. This was about 1967 and even home video not yet available. I made sever such conceptual drawings and not of what I thought would have been neat things to have. Most of what I designed was highly impractical at the time, but the ideas looked good on paper. A good idea is to also date your notes and drawings since you will be able to revisit them later.
As to the area of droids. I really believe that a droid is an advanced robot. The robots that are available to individuals today are far too expensive, flimsy, and impractical to be called droids. The people that make droids are doing this for ascetic reasons. The people that build robots are doing so for research (learning) purposes. We need to begin to merge these to areas to come to a more common goal. A list of features and capabilities of common robot capabilities need to created. Yes we all have built little track drive robots. many of us have done the sonar, or infrared ranging thing. Each of the robotic hobbyist have done things that others have done. This is no way to move into the droid era.
If the computer industry is to be used as an example look at the evolutionary steps. The first real PC started out very crude. We began with a computer with two floppy disk, a monochrome display and a keyboard. Printers were textual only. Since there was no graphics there was no need for color monitors or mice. When Windows was accepted, we needed color graphic boards, color monitors and mice to point and select. With more and more applications being developed everyday larger disk drives were needed as well as larger RAM memories. Modems were still slow and were not very popular since there was no one to really talk to. When the internet came this changed very rapidly. Modems became very fast and moved inside the computer. Some homes began to have more than one computer and the idea of either net became popular. As the years went by many things have became standard and even expected on a modern computer. I can remember a time some would question why anyone would need a computer to now when most people have them.
Droids can be considered the same way. We begin with a basic droid that can do some elementary functions, yet designed. As the rears progress these functions will be advanced and new functions will be added. It is a given that the droid should contain a fully functional compute connected to both telephone and to the internet, or what ever its equivalent is at the time. All popular multi media and entertainment such as gaming should also be present. The system should be completely non-volatile. This means that everything is completely backed up including configuration to a safe location should the droid be damages, lost or stolen. A personality is an absolute must. People like to bond to things that appeal to them. This should be just as true with a personal droid. Since it is mobile it is assumed that it will be with its owner. If another user is allowed to access the droid it should have a way to upload that user's profiles to work with the other person. PCs on networks already can do some of this but on a limited scale.
Something that I would like to see is hack-ability. The computer Gods of today have literally stolen our computers from us. They have put locks in the computes so that the next generation hackers can not develop things inside the computer. I an a hardware hacker from the beginning of personal computers. I really resent that fact that all of the computer development languages have locked out the ability to easily talk to the hardware that is attached. One of these computer Gods tries to rationalize the locking of the user from the hardware as being some sort of protection from crashing the computer because programmers are stupid. If so then give us some hardware that is separate from the secure stuff. Give us some access to some of the ports and perhaps our own buss or slot to connect to.
At any rate droids should be designed from the very beginning to be hacked. This allows and encourages future invention of future droids and add-ons. Legislate the openness of systems like this. Make it illegal to prevent upgradeability and hacking into any part of the droid. Is security is a problem than lock down the items that need to be secure and open those that do not.
We need to generate new cultural ideals that discourage the prejudice of new technologies. There are a lot of people that go out of there way to discourage new things like, the light bulb, telephone, automobiles, motion pictures, record players, computers, CD, VHS /DVD and much more. I know people that need access to the internet and are still afraid to use the computer. We need to attack this a another technical problem. The kids that I grew up with never knew of computers when they were young. The kids today have fully accepted computers ad being a natural part of their environment. This is part to the fact that schools have incorporated computers into their classrooms as fixtures like the chalkboard. When the principle of my high school retired the ran for and got a position on our local school board. When an item came up to incorporate computers into the classrooms he filibustered ageist them. He believed that they were only good for playing games and that the students would not learn anything. He said that they schools give all of the students hand chalkboards and chalk and forget about taking this path. He and others with similar prejudice successfully delayed for years our district from moving up to modern times. This is type of people that will continually block us from moving to the next technological level of evolution.
Now we need to begin somewhere and primitive robots are all we have. So let us start there. To build a droid we need a drive platform. This platform need to drive on the surface that we choose. In my experiments I will be using several surfaces. When weather permits I wild like to be outside and use cement. It is broken and rough. Other time I will be in the house where I have hardwood floors. My shop is in the basement where there is once again cement that is smooth and cleaner. I have a dog that leaves hair that can be picked up in the drive and wheel mechanisms. So it is expected that at some point the motor axels an the caster will jam up with hair and dirt.
Well you get the idea. What I looking for as an outcome of this work is a robot platform that can be used to do experiments on long after the initial construction is done. The smaller platforms are too small to allow much in the way of additions and the larger platforms are too large to be used in the house.
Next generation platform robot:
This generation has been constructed from aluminum box tubing. My last large robot was constructed from a 0.5" plywood base. Two eighteen inch disk cut from this plywood. Four 1/4 - 20 rods were used to space the two disk apart using washers and nuts. The lower deck has a aluminum channel mounted to to support a shaft that connects to a fork that suspends the drive wheel as well as being the steering wheel. There are two brackets that support the rear axle that has a wheel on each side. This robot looks like tricycle with two disk platforms. Power was supplied from a 12 volt lead acid motorcycle battery. Testing was done using a two channel radio control unit and two power controllers.
I was working on an autonomous system, but through the lack of high powered P-Channel N-Channel FET combinations I was limited to relay drive or something quite complex to drive the motors. My father got sick at the tem and died and I just shelved this robot. I continued by using smaller robots (6 inches or less). The problem is the severe limitations to small robots. Over the year I considered building a larger robot, but lacked the enthusiasm. Also there has still been the problem of finding parts like gear head motors and pre-made H-drivers. For some reason pre-made H-bridge drivers have been scarce. Those that have been available had been expensive and complicated. The same is true with gear head motors. Recently with the popularity of robot completions many of these parts are becoming more common as well as the prices becoming lower.
My interest is in autonomous robotics with a minimum amount of control. I would like to setup an area outside or in the basement to conduct experiments. It would be nice to experiment with auto navigation and object recognition. There are a lot of off the shelf components that will help with this. My most recent work that I have done in this area comes from my Lego Mindstorms system. I also have several of the little self-contained sonar boards. These work great. At any rate since I now have a larger platform for experimentation I can work with all kinds of sensors and cameras. I would also like be able to pick up samples and do other things. My interest is in creating probe robots and not killer robots and I have traded speed for accuracy and power.
The design of this robot was a bit cleaner that usual for projects of this complexity. There are always parts that are hard to find. Thing like small computers are off the shelf, while motors, wheels and drive electronics are still specialty items. The following is a list of some of the off the shelf items I have used.
The frame and mechanical is custom. The idea is to use all stock electronics and focus on the mechanical build up. The frame is made of 0.75" by 0.75" box tubing with a 1/16" wall. I cut two 12" length and two 10.5 inch pieces. I used a 14" chop saw to cut the tubing. After de-burring the tubing I cleaned a half inch from the end on one piece and around the end of the other to allow a good solder connection. I tried to get the solder to flow and made a mess. The solder only stuck to the edged what it had been cut. The solder just peddled up and did not flow. I soon realized that this tubing had been anodized to protect the surface. I used a grinder with a sanding drum and removed the anodized surface at where it needed to be soldered. This time the solder flowed. I have constructed the entire frame from aluminum. The solder is Alumaloy and it works great with just a propane torch for a heat source. The connections are good and strong. I made the motor brackets from aluminum stock and machined them to work with my motors. I then soldered them to the frame. The battery bracket will also be soldered is as well as the electronics plate and the antenna bracket. This makes for a lighter and stronger frame than with wood and lighter that steel.
Pictures will come soon.
I am going to start by adding non-contact sensors for simple navigation. this will at least allow me to get back up to a minimal level. I am really getting tired of having to direct link my robots and controllers. I purchased a pair of Lemos RS232 wireless Bluetooth interfaces.
They allow RS232 communication up about 100 Meters. Most robotics apps are usually line of sight and can use less. If you communicate in either hyper-term or visual basic you can tell the robot what to do form a computer. My on board computer is a X186 and has a dos operating system. This means that its communications is naturally serial. If you need to get fancy and be graphical you can write control panels in Visual Basic assuming you have the MScomm32.ocx loaded. This allows Visual Basic to communicate in serial to other devices. On my robot I am soldering a support to mount the Lemos device. I am also adding a PIC controller that will monitor this serial port for certain communications strings such as reset and shutdown commands. During operation it will also look for watchdog commands. This is done to ensure that communications is not lost and the robot runs off on its own.
This robot's software is written in Borland C. It is simply, easy to use and quite reliable. Unlike Visual Basic or any of the other PC languages available today Borland C still has its serial IO commands. My ultimate direction is to create a a standard graphic language. I wish to use graphical blocks to indicate chunks of code like read an a2d port. My language has been evolving and I think am close to a solution.
What will happen is that a user can select function blocks from a library. Each selected block is then dropped on a sheet that represents program flow. Data is linked internal to each block. Each piece of code can be property edited or code edited. If the property is edited that the programmer can change things like variable names or port addresses without reworking the code. If the programmer chooses he or she can also look at the code or even change it. Similarly the programmer can also create new code blocks. Once a block works the programmer need never rewrite this function again.
