Wireless Controll Advice
nondem
Posts: 15
I'm building a remote controlled boe-bot that will carry a battery powered web-cam. Kind of a mars rover thing, that I could drive around my property and look around via the web from work. I'll be developing a web interface for it in Visual Basic.Net The webcam is self-contained with it's own 802.11 webserver so that's handled.
I'm trying to decide which method of communication to use to send movement commands from my laptop to the bot. I needs to have it's own communication path with a controlling "server" which will consist of my laptop.
I've been looking at the 900mhz transmitter/receivers, the one-way transmitter > receiver combo, and the bluetooth module that Parallax offers. I also have been looking at using a generic bluetooth serial adapter on the bot. The bluetooth option is attractive because it would'nt require any type of custom hardware at the laptop end - where the non-bluetooth options would all require a custom solution at the laptop end to drive the non-bluetooth hardware.
So anyway - I plan on sending commands to the bot from as far as 200 feet and possibly through some walls and such and I'm concerned about the bluetooths range limits, ... I would appreciate any input from you guys on what would be the best path to take.
I'm trying to decide which method of communication to use to send movement commands from my laptop to the bot. I needs to have it's own communication path with a controlling "server" which will consist of my laptop.
I've been looking at the 900mhz transmitter/receivers, the one-way transmitter > receiver combo, and the bluetooth module that Parallax offers. I also have been looking at using a generic bluetooth serial adapter on the bot. The bluetooth option is attractive because it would'nt require any type of custom hardware at the laptop end - where the non-bluetooth options would all require a custom solution at the laptop end to drive the non-bluetooth hardware.
So anyway - I plan on sending commands to the bot from as far as 200 feet and possibly through some walls and such and I'm concerned about the bluetooths range limits, ... I would appreciate any input from you guys on what would be the best path to take.
Comments
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
~~ dRu ~~
You said your robot would roam about your property. Does that mean it is an outside machine with plenty of power?
I built a skid-steer robot named Ugly Buster. Although I have not done it yet, before long I will simply mount a tablet PC on him and solve all my remote communications problems via ethernet. I will run the tablet PC off its own battery and Ugly Buster will use his own two battery packs (12vdc, 24aH and 7.2vdc, 8400maH; grounds tied together).
Like most of us, I imagine, my entire house is wireless and Ugly Buster will get along well indoors with the tablet PC. I THINK he will do well in my back yard. If not, I will stick an access point outside. Currently, the decision-making is being done by Stamps . . . anywhere from one to five on board. I've only used the BS2p40 "master" and a single BS2 "slave" once. The other BS2 slaves are idle, awaiting my programming expertise. Ha! Not really! They are waiting on me having some TIME to do it!
Usually, for demo and development purposes, I run the whole thing off a single BS2. (This was predicted by several on this forum during my design phase.)
Wouldn't a wireless ethernet module be nice? Basically, that is my purpose in mounting a tablet PC on my robot, although I will certainly use the compute power and the disk storage, as well.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
The goal is to have a "rover" bot that can live alone in the outdoors. I'm building a chassis that can support a 12v 7ah lead/acid battery.
I can regulate it down to 5-6vdc and supply everything from it. I plan to use a solar battery charger to keep it charged without need a "docking-station". I'm thinking an automotive solar battery charger could serve as a weather-roof against rain and also at least extend the time it could run w/out needing a proper charger. I envision building some logic into it that would make it sleep if the battery got low and occasionally wake-up and see if it has enought power for some type of operations.
I even considered using a photoresistor to help it seek out a sunny spot to hang-out and charge....All these grand-plans are long term stuff. For now - I just wanna send commands to it here around the house and prove the concept.
Until I get the wireless control part worked out I'm writing the code in VB.NET to send commands to the bot using the SERIN command via a plain wire plugged into the port on the BOE. Thats not going to well though - I can get the serout command to send me text into Hyperterminal proving the connection config is right - but I can't get the bot to respond to commands sent to the same port from my application when using a command like : SERIN 2, 16468, [noparse][[/noparse]WAIT("fwd")] - It just sits there like it never got the "fwd" string, but I did confirm it was sent from my VB.NET app using a loopback comport emulator(com0com). I've been eyeballing the command reference ... I haven't given up [noparse]:)[/noparse]
I have a lot of experience with networking stuff. (Mike has a lot more.) At one point, I thought ARCnet would rule the world. I thought it would be released in faster and faster versions. I thought that a deterministic protocol was the future of the world and that a non-deterministic protocol--say, like ETHERNET (in the beginning of my "time", Aloha Net)--would go the way of the technological Dodo bird--the 8088 based PC (which was to come years later). After all, Ethernet is designed to communicate by recovering from failure. What kind of deal is that?
But, I was wrong about all of the above. Datapoint never released a faster ARCnet. The 8088 was rapidly replaced by the 80286. (Gee! I even had some 80186 stuff! That was fun!) And, the hurt to end all hurts was when I finally came to understand that recovering from failure is the true path to success in all endeavors, be they human or otherwise. Ethernet Rules. Translation: You simply cannot combat the purchasing power of the masses. (As much as we like or dislike him, Mr. Gates realized this long ago.)
In the meantime, there are other attractive things like Bluetooth, etc. All of them fall short of what all of us desire: We want to treat our robots as network devices. Nothing could be simpler. One of these days, an Ethernet manufacturer is going to see a market in what we do and love; hobby robotics. When that happens, we will succeed through recovery from failure. ip is here to stay. One day, every "thing" that uses electrical power will have an ip address for control and status checking. Do you doubt it? If so, think of Internet II. Not the speed, but the expanded ip address range!
Until then, I will use a tablet PC on my machine to emulate the real Ethernet controller I covet. In the meantime, the 'bot-local compute power will be nice, as well as the storage. I still dream of the tablet PC talking across the Ethernet to a rack of 1u machines, each dedicated to a single function of the Stamp controlled-tablet PC augmented robot.
Further and on a different topic, the Propeller, I am afraid, will be relagated to less than stellar success. It is a FANTASTIC concept implemented in a most excellent manner. The trouble is--as we would say in the old days--"it don't boot DOS." Translated, this means it does not run a version of the most popular language on PLANET EARTH: BASIC. It makes NO DIFFERENCE if Spin or Propeller assembly is better or not. The fact is that from a MARKETING standpoint--which is the ONLY standpoint that counts if you are interested in product success--that the super-processor you produce must run some version of BASIC to be a success with the MASSES. If your marketing strategy does not include the masses (and mass volume), then you can afford to produce something like the Propeller.
This is a very fundamental point. Not even C-based stuff has the mass appeal of BASIC. There is absolutely no comparison. The minds are many that can immediately comprehend BASIC and immediately cut code and immediately improve upon it. It is akin to a Darwinian process. The people that can learn BASIC in a few days or weeks are LEGION. BASIC will go down in history as the most successful language ever developed, just like the lowly .22lr cartridge, which is over a century old.
From my point of view, my point of view is odd: I "started life" as an assembler coder and an ARCnet expert. I built the very first useful ARCnet systems in the academic world and they served exceptionally in a campus-wide environment for both student and (light) administrative use. But, Ethernet came to rule the world of networking and BASIC to rule the world of application software.
The buying power of the masses made this happen. It certainly was not the superiority of Ethernet and BASIC.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Unfortuneately, it is because I am old and have many experiences, most of them difficult. However, the difficult ones are the most valuable.
Thank you.
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
It depends on what you want. Bluetooth can go up to 100 meters if you use class 1 devices. This can get expensive, but it does give you more flexibility in that you can use pretty much any bluetooth enabled computer for testing.
Bill,
I don't understand the argument that ethernet is required for IP addressing. There are many different layer 1 interfaces that allow for IP addressing. In fact, you can transport IP packets over serial, parallel, usb, ethernet, wifi, wimax, bluetooth, etc.
I think the reason why ethernet isn't as widely available as you would like is because of the complexities of embedded implementation. Most people don't want to mess with sockets level stuff. They want simple encapsulated serial port profiles provided by Bluetooth and device servers.
There are plenty of companies that sell 802.11 wifi device servers that would alleviate the need for you to use your expensive tablet PC on your robot. It will cost more than $100, but it will definitely be better than having to mess with a PC and writing bridging software.
Harrison
Post Edited (Harrison.) : 12/2/2007 9:01:22 AM GMT
Of course, you are right. You can do ip addressing with smoke signals, if necessary! [noparse]:)[/noparse]
My whole point is that to address mass markets and the uncountable discretionary spending power they represent, you must make your products attractive to those markets. To date, in the history of modern computing (gee, I wonder what the "history of modern computing means!") the two common denominators attracting people have been Ethernet and BASIC.
I am not saying they are the "best" solutions to the world's hobby robotics problems. I am saying they would doubtless be the most popular solutions to the masses. That's where the money is.
Somehow, despite the fact that I cannot talk to my skid-steer, Ugly Buster, via Ethernet, I will certainly enjoy further development of him today. By the way, having that tablet-PC mounted on him will allow me to RDP in and program my Stamps from afar. Perhaps, there are other solutions to even this, but this one is sure tried and true and many, many people on this forum have tried to figure out how to do exactly that. I am not·a fanatic about this and I LOVE the Parallax products! (I wish they would build a module version of the SX, though. [noparse]:)[/noparse] )
Merry Christmas!
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Harrison help me make a really cool Web interface (he did the hard part) that will dynamically display sensor input and also allow for control of the bot. I haven't tested the the range on the transmitters but. the data sheets says 500ft+ this is my communication path: PC <--> WWW <--> PINK <--> RF module <--> RF Module <--> Propeller.
Good Luck!
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"A complex design is the sign of an inferior designer." - Jamie Hyneman, Myth Buster
DGSwaner
What I am saying--and I am sure you understand--is that it should not be that difficult. We need something to stick on the 'bot and "plugs into" a Stamp and immediately the Stamp is available across your wireless Internet in both programming and data transfer modes.
I bet you are having fun with that web interface, though! [noparse]:)[/noparse]
--Bill
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
I totally agree with you on the last post you made. Just so you know where I am coming from (the younger generation) I'm a 19 year old Biology/History/Psychology Senior currently enrolled in University. I have had no time to learn how to properly navigate and manipulate linux and I'm more of a "hobbiest" then an engineer. Over the years I have taken 3 programming classes, C and Java (dropped Java). I know my limits and knowing that I'm not a good coder is one of them. The Basic stamp has alleviated alot of my coding anxieties in that it is straight forward. As a summer project I converted an entire 114 acre school campus to wifi using 10 Buffalo WHR-HP-G54 routers ($55 each). Now all I want to do is to build wireless applications that allow me to remotely control or monitor certain aspects. For instance, I want to make the time clock networkable, put a gas pump on the network, and open gates remotely. The circuity is the easy part, but when it comes to putting it on the network I find myself having to purchase a PINK and connect that to a wireless router and then merge that wireless signal with that of the schools. As you can imagine that puts me at around $200 for such a simple task. If Parallax developed this "chip" that I could just plug into the BS2 breadboard.... the wonders...
Are there any other workarounds that I can use? I am not an engineer, I'm a college kid, all I want to do is to economically build robots that help people do every day tasks but am being limited from the transition of wired to wireless.
Merry Christmas All
-Luke
The only issue I see with using Lantronix solutions is that they don't provide nearly as much information as other manufacturers. There is extremely limited documentation on how to use the internal webservers for their products (I believe you have to use TFTP and some other non authenticated protocol).
www.gridconnect.com/matchportar.html
I'm also a college kid so I understand where you are coming from. I definitely go for the most inexpensive route (hence why I am staying with wired ethernet till everything is wireless...) I develop my own tcp/ip stacks just so I can use inexpensive mac/phy chips to reduce the cost of hardware.
Off-topic: Did you happen to use third party firmware (such as OpenWRT or DD-WRT) on the Buffalo routers? I really want to get one to see how well they work compared to Linksys, unfortunately it's nearly impossible to buy one due to the whole '802.11a/g infringes on Australian patents' thing.
Yes, I think you do understand it perfectly and your circumstances also perfectly illustrate my point. Communications of this nature should not be technical. There should be a device that we simply plug in and it works.
Now, some will say that if you do not have the expertise to build a proper communications link, then perhaps you should do something else for a living or a hobby. Such an attitude is TOTALLY WRONG. It is not your electronics or computer expertise that is important. It is your MIND. (Your example shows this very well.)
My point is that the more minds we have that can concentrate on robotics-type development and automation, the better off we shall find ourselves and the faster the technology will develop. It is a cycle that never ends.
I am not a college student. Plus, I do not have a lot of electronics background. But, I am/was a college dean (retired October 1!) and Chief Information Officer. I have well over 30 years watching these cycles operate and enabling THINKERS with technology. It is their mind I am after, not their technology expertise. The technology should serve THEM. If it does, then they are free to CREATE.
Parallax is probably the wrong company to build this product, anyway. It must be a marketing success in order to have any longevity. I am not sure such a product would have the right "fit" in the Parallax line up. Then again, it might!
--Bill; on his soapbox
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You are what you write.
Just wanted to give you some input on your question.
Buffalo + DD is much better performing than Buffalo with standard firmware.· The power output options and additional routing control are well worth the flash.
If you're looking for range, Buffalo + DD or Linky + Tomato are awesome options.