Help designing a circuit board

[juggler325]

I'm a senior in High School and am looking to design a device that will assist visually impaired dogs by emitting a noise when they near solid objects. First, however, I need an inexpensive sensor that can tell the presence of walls, doors, couches, etc., from about a foot away. I have experimented with a cad cell, but am wondering if there are better alternatives. My budget is pretty large but since I'm designing the product for common use it would be nice to find cheaper parts. Do you have any suggestions of what I could use? Thank you for your help.

[billchapman]

What a cool problem! Here's a website I think you'll find helpful:

http://fargo.itp.tsoa.nyu.edu/~kb561/pc ... inder.html

There's a nice brief discussion about how rangefinders work as well as a few pictures, model numbers, prices, etc. The sonar device would probably drive dogs nuts (and maybe dog owners too) but the infrared device might work for you. You can find further information on the performance of one of these IR sensors as tested by some Carnegie Mellon students here:

http://www-2.cs.cmu.edu/~pprk/tech_info.html

Good luck, I have 3 dogs--maybe someday I'll need your invention!

[Juggler325]

Thank you so much for your help! The infrared ranger is exactly the device I was looking for. As you can probably tell, I have very little experience with electronics but I'm hoping the research and development of this project will help fix that. Depending upon how complicated I design the project there are multiple issues that could arise. Already I am predicting a few problems such as the fact that when the dog sits, the sensor may trigger on the floor and produce constant noise. Also, I'm not sure of where to place the device on the dog where it could give a consistent forward reading without sometimes being blocked by the movements of the dog's head. Another suggestion someone has given me is to substitute a vibrator(like the ones on cell phones) as the warning, instead of an audible sound. If you have any thoughts on any of these ideas or how I could put it all into a circuit, I would greatly appreciate your input. Thank you.

[billchapman]

Good points, you are clearly thinking ahead very well! I’d suggest you keep a list of these specific problems as you proceed developing the basic concept and technology. Periodically revisit your basic idea with the new considerations. For example, you might think about collision avoidance vs. proximity detection. What’s the difference? Well, in collision-avoidance mode, your circuit could monitor the proximity and the rate of change of proximity. Can you see how you could combine these to solve some of the problems you’ve identified?

On developing the circuit, I can suggest an approach: Think up some algorithms that might work. When you have the problem reduced to an algorithm, you will know just what you expect the electronics to do. Then you can search for implementations in electronics textbooks or Op-amp “cookbooks�. Also search the websites of the vendors of the IR rangefinders for application notes and “white papers� --these often include sample circuits.

When we are developing new technology for products, my team does just what you are doing: We work on the core technology, while considering what ancillary problems we’ll need to solve to make the product real. We make presentations to our top management and to customers showing what we’ve achieved along with our best estimate of how to solve the remaining problems and what it will cost in time, effort and money. In the corporate world, we then decide if proceeding with the development is a good investment of our resources for the company and its stockholders.

I mention all this to point out that a demonstration of the core technology combined with a well-considered assessment of how to develop it into the final gizmo forms, by itself, a valuable “product�.

One more recommendation: Keep good notes and sketches on your ideas in blue or black ink in a bound notebook with numbered pages. Date and initial all of your entries. This type of documentation forms a basis for patent disclosures!

[juggler325]

Thankfully I have been following some of your advice. I have begun to write down some of the potential problems and since my project involves both electrical and animal elements, the issues can be wide ranging. Pretty soon I'm hoping to meet with a veterinarian and discuss some of the canine problems. Observing my aunt's blind dog, I noticed that he usually had his head down when walking; probably to take advantage of his sense of smell. If this is consistent I may be able to put the sensor on the top of the collar where it could take readings when the dog's head is down. I want to talk with the veterinarian to see if most visually impaired dogs have similar habits.
As a result of Thanksgiving, I have had a few extra days off and so I checked out some books from the library related to electronics and have been studying the basics for the past few days. Much of the information is a bit overwhelming and some information is simply not available through the library, but luckily I do have access to some other resources. A technical school adjoining my high school offers a computer servicing course and the teacher has already been a great help in getting this project started. Once school resumes, I will ask him about some op-amp "cookbooks" and electronic textbooks. I already have borrowed a few books from him to start gaining a better understanding. I truly am very inexperienced when it comes to electronics so the teacher has been a great help in showing me the basics. I appretiate your patience in working with me as well! Some of your recommendations are a little hard for me to understand because of my limited knowledge, but on monday I am going to visit the teacher and ask him to clarify some of it to me. If I still have questions, I'll probably post them here.
I have looked into the infrared ranger and how to modify it to my needs but I have a lot more research to do. Almost all of it is online because I haven't been able to find many books that deal with it. I am planning on ordering one after I consult with the teacher at school. I really need to see it and test some of its characteristics to understand how it works and get a feel for some of the problems or unexpected benefits that it's limitations could cause. I like the fact that part#: R21-IR05 has an adjustable range threshold which might make my testing easier.
Following the reccomendations of teachers and websites, I have been keeping a day to day log of my research and activities. My penmanship is poor, however, so I have been typing the log and anything else I have needed. I have also begun a works cited of the resources I have been using. I see your point in documenting with pen and paper. My hand writing is legible, just messy, and I am very concerned about neatness. This process has worked out so far because I bring a disk whenever I research and add to my log as I go along. Would you reccomend that I begin writing them instead so I have more firm proof and documentation of my progress?
Thank you again for your advice, it has been a real help!

[Guest]

No need to write by hand, here is what you do: Print out any computer-generated output you want including text, graphics, plots, spreadsheets, etc. Tape three edges of each printed page into your book (reduce and trim to fit if necessary). Initial or sign and date in ink across the remaining edge so that part of your writing is on the notebook page and part is on the added sheet. Then tape over the signed edge (tranparent tape of course).

Other notebook tips: Don't remove any pages. If you accidently skip a page, draw a diagonal line in ink through it. Cross out mistakes with a single line so that the material is still legible. These habits are well-appreciated by research advisors in college and grad schools and are essential in government and corporate laboratories. As previously mentioned, they also preserve the value of your notebook as a legal document for patent purposes.

On undestanding my recommendations: I don't mean to be obscure but I also don't want to rob you of any "a-ha" experiences so I may start off a little vague. Here's what I was getting at: The IR ranger outputs a voltage related to proximity. Proximity changing in time (e.g., dog approaching obstacle) will thus result in output voltage chaging in time. Try making a table for different "states of the dog": Dog sitting, dog walking, dog running, toward an object, away from an object, object close, object far etc. What is the voltage in each case? How is the voltage changing in time in each case (is it increasing? decreasing? how fast?) Now you have a table of dog-states as represented by the ranger output. Which states do you want to alarm?

The voltage response of the ranger is not linear. This is a detail that can be handled later but here's a teaser: There's a way in which this non-linearity can benefit your application!

P.S. Great observation and insight on the canine behavior!

[juggler325]

You really haven’t been too obscure for me. Every time you mention a certain element that I might want to look into, I try to research it online. I’m actually a bit concerned that I might lean too heavily on outside resources and not on my own studies and understanding. After all, a science project isn’t any good if you can’t explain the science behind it! I just want to make sure I’ve checked out all my resources before I ask for clarification.

When you mentioned about an algorithm, were you thinking about a flow chart? A flow chart would be a great way in which to visually show my thought process, from conceptual to reality. Thanks for the recommendation.

At this point I have to order parts, look at more web sites, and start refining my procedure and the actual experiment. A lot of the information that I need, I’m hoping to find through my school. Until I talk with some teachers and get a few more books, my questions for you would probably be pretty trivial. With that in mind, it might be a few days before I have anything important enough to post.

Thank you once again!

[Juggler325]

Well, I guess my "few days" turned out to be a bit longer. I almost gave up on my idea because of the cost and questions involved but I decided to still give it a shot.

I retained the same general concept but have included a few new parts. To sense approaching objects, I'm going to try using a MS-401 PIR module in conjuction with an IR LED. I bought it at:
http://www.hobby-boards.com

To detect the motion of the dog, I'm going to try using part of a pedometer. Hopefully it will signal when the dog is at rest and when it is moving.

To actually alert the dog, I will either connect a piezo buzzer or a vibrator like those on cell phones.

To tie it all together I bought a basic stamp kit from Radio Shack. It uses the "homework Board" and I have been experimenting with how to program it correctly.

When all put together, The sensor and LED will probably be on an antenna hooked to a harness on the dog.

I just wanted to let you know the current status and progress. I have yet to make a functioning prototype but I'm still working on it. I've just finished assembling the PIR mini motion detector so I'm still unsure if it will do what I'm looking for.

If you have any other suggestions I would be happy to hear them. Obviously I haven't been very good about continuing to update and check this post but I will certainly try to do better in the future. Thanks!

[deleted-2574]

Well, I guess my "few days" turned out to be a bit longer. I almost gave up on my idea because of the cost and questions involved but I decided to still give it a shot.
...

Way to go! Sounds like you have the perseverance to make a scientist!

[juggler325]

Thank's for the encouragement : ). I'm glad I continued; I am making progress. If I do finally get it fully functional, I'll post it here. So far I have a unit that will detect motion. My hope was to have an Infrared LED shine off obstables and have the PIR sensor trigger on the reflected light. It looks like it may be to difficult to create enough reflective IR light so I think I'll go back to the original plan and use an infrared ranger. At least I know now what won't work, now I just have to find what will!

[deleted-2574]

Dear juggler325,

Sounds great! Let us know how it's going, and if you have any other questions or need any other help.

[rrainville]

This week I have off from school so I've stepped up the research and parts collecting.

A local cell phone dealer was kind enough to let me take two mini vibrators from some of his old phones. Their compact size should be perfect for placing on the dog and just one provides a very definite signal. I also have a small piezo buzzer that I would like to try out too.

The other day I received the Sharp GP2DO5 in the mail and it's even smaller than I expected. Hopefully there should be no problem mounting it on a small antenna on the dog's back. It looks easy enough to interface it with the basic stamp 2 though I haven't tried it yet.

I think my most challenging part will be getting the pedometer to do what I want. I don't want the device consuming energy when the dog is resting so my plan was to use a pedometer to show the dog's movement. If it gets up, the pedometer should trigger and send out a signal. Do you have any other ideas of ways to detect the dog's motion?

All in all things have been going successfully and the combined costs of the components hovers around $50, an amount that would fit my design criteria. If you have any suggestions or concerns I would be happy to hear them. Thanks!

[EDS]

Hi rrainville,

Sounds like a neat project.

The pedometer sounds like a good idea. You might want to add a timer that keeps the device going for a few minutes in case the pedometer only triggers occasionally.

If the pedometer doesn't work, you could also try using a commercial accelerometer. Among other options, Analog Devices makes a line of single IC package, low-g accelerometers that sell for something like $5-15.

There are also devices generally used as low accuracy shock triggers that consist of a cm long strip made from layers of flexible, conductive plastic with a tiny weight on one end and two leads coming out of the other end. When they vibrate, they bend and their resistance changes. They sell for around $1 each, although finding small quantities may be tough. I know someone who has had a lot of experience using them, and will ask him for some some part numbers and vendors in case you're interested. (I can't remember what they're called at the moment.)

Good luck,
Erik

[deleted-2574]

Hi Erik,

Re your recent posts (Re: Collecting parts for an object detector and Re: How to use Standard Deviation results), great posts!

Then I noticed the time-stamps of the postings, way early AM. Is this typical? I know this is not Science Buddies related, except as it concerns working scientists and science students. My eldest son plans to be a physics grad. student in the fall.

[rrainville]

Thanks Eric for your advice, I'll definitly look into some of those options.

Sorry, I realize there might be a bit of confusion, I earlier posted as Juggler325, same person though. I'll keep you posted on my progress. Thanks again for your assistance

[EDS]

rrainville -

Hi again,

I looked into those cheap plastic vibration/acceleration sensors. I was wrong about how they work - they actually use a piezoelectric plastic that generates a current whenever it gets squeezed, rather than a changing resistance.

But, I was right that they cost about a buck and are easy to use. The material is called Polyvinylidene Fluoride or PVDF, and is often sold under the trade name Kynar. If you do want to try them and you find it impossible to buy small quantities, let me know.

I'm looking forward to hearing how your project works out. I hope you'll stop by and give us a run-down when it's finished.

davidkallman -

Thanks David. I'm always a little afraid that I'm stepping on other people's toes when I answer posts in existing threads, but I don't seem to have offended anyone too badly yet.

Congratulations to your son. I imagine it must be getting close to prospective visiting time. I found the applications and the choosing a school part utterly terrifying, but everything since has been great.

I tend to be a night person by nature. Whenever I don't need to interact with other people for a couple days I usually slip into a dawn-to-noon sleep schedule. Meetings and the schedules of the people with whom I work keep me from swinging to quite that extreme most of the time.

I wouldn't say it's particularly common in the field, although there's often a great deal of schedule flexibility, at least after the first year or two of classes is over. Students and faculty tend to set their own schedules. Most people in my group show up around 11am and stay until 7-9 in the evening, although there are some exceptions. I'm rarely the only person in the building at 1:00 AM.

One thing that does seem fairly universal, at least in experimental groups, is a tendency to work in spurts - 70 hours one week, and half days the next.

take care,
Erik

[rrainville]

Thankfully I am finally able to access this sight easily again. My computer died about 2 months ago and then my new one was made inoperable by a power surge so now I'm on computer number 3.

I continued to improve my project until I was ready to test it on a real blind dog. With the permission of its owners, I tried the device on a twelve year old golden retriever who had been blind from cataracts for 12 years. Though I didn't leave it on the dog long enough to let it learn (through classical conditioning) how it worked, the testing gave me the opportunity see the many ways in which the project could be improved. Though I am a high school senior and cannot participate in Sci-fair next year I still want to continue the project and hopefully reduce its size and improve its functioning so that it can possibly become a marketable product. Right now the device attaches to the top of a harness and I want to modify it to fit on a collar.

I did get the opportunity to attend Intel ISEF this year in Phoenix Arizona as one of two Vermont finalists. I am in the finalist directory under Ryan Rainville, EN122 (last year I was EN116). Though I didn't win any awards the trip itself was amazing and helped me to generate new ideas that I hope to work on in college.

Next year I will be attending Wentworth Institute in Boston and pursuing a major in electromechanical engineering and I know that this project has only helped to enforce my interests in engineering.

Thank you to everyone who has helped me to develop this project. The assistance provided through this website has been a great source of help in working through the obstacles and questions I encountered. As I continue to work on improving the design I'm sure new issues will surface and hopefully I can keep updating you on my progress.

Congratulations also to the Science Buddies mentors who participated in the Intel ISEF.

Thanks again,

rrainville

[EDS]

Hi Ryan,

Thanks for the update - I was curious how things worked out. Feel free to drop by and say, "hi" again in the future.

Glad to hear you ended up with a working prototype. Sounds like an all around neat project.

Congratulations on going to the ISEF. Just getting there is very cool, awards or not.

Have fun in college! I expect you will. The transition from doing all the fun stuff *after* school to being able to do fun stuff *during* school can be a great joy.

Best,
Erik

[deleted-2574]

Hi Ryan!

Thanks for the update, and I second Erik's comments.

Re marketing your device: my expertise is in software marketing, particularly programming languages, i.e. compilers. So, I'm not too familiar with marketing your type of device.

For your device, I think a lot of other issues come up, e.g., patents. It may be best to consult someone who has some expertise in marketing the device you're looking at before you go much further. Maybe the local Better Business Bureau can help.

It is possible that they suggest selling the rights to the patent to a third party. Then, much less capital is required, though, upside gains will be limited.

Best of luck in college!