Hello all,
I am currently working on a final year project about a seaperch ROV (remotely operated vehicle) which i want to convert it into an AUV (Autonomous underwater vehicle). I have trouble here considering the lack of knowledge in both electrical and programming stuff since the diploma course that i am currently taking now only teaches us naval architecture.
Ultimately, i would like to incorporate the following into my AUV :
- MPX5050DP IC pressure sensor - to gauge depth
- LSM303 LSM303DLHC 3-Axis Electronic Compass Acceleration Module
- Arduino Uno
Currently my Seaperch ROV is running on 3 x 12V DC motors.
To sum it up, i need some general advice on the arduino coding as well as i would like to enquire if i need anything extra, eg. datalogger (for readings), etc for this project. I would appreciate any suggestions for this project.
I would be grateful if you would take some time off to help me with my project. 
Thank you.
You are going after a big ambitious project. I don't know anything about underwater nav, but if your vehicle is to be autonomous, it needs to arrive somewhere definite at end-of-mission? On land, there are no currents, and underwater, they will complicate navigation.
I assume you do not want to do the math involved with the accelerometer, which can be very sophisitcated (trust me, my Ph.D. is in math), so be sure that whatever accelerometer you use provides you with software support for location calculation. And of course, that SW must be compatible with your Uno. It also seems likely to me that an UNO is not a powerful enough chip for your task. I'd suggest using something bigger, say a Mega or a Teensy. You may need to be able to support hardware floating point in a separate chip. Big job, but I like ambitious people. Good luck.
You cannot use an accelerometer to calculate your position. Well, you can for a very short period, but errors in your position calculation will add up very quickly and you'll have lost any hope of knowing where you are.
For data logging you can use an SD card. You should have no problems finding plenty examples of this.
First thought: "what will the unit actually do autonomously?"
jrdoner asked this also ...
What is the task?
jrdoner:
You are going after a big ambitious project. I don't know anything about underwater nav, but if your vehicle is to be autonomous, it needs to arrive somewhere definite at end-of-mission? On land, there are no currents, and underwater, they will complicate navigation.
I assume you do not want to do the math involved with the accelerometer, which can be very sophisitcated (trust me, my Ph.D. is in math), so be sure that whatever accelerometer you use provides you with software support for location calculation. And of course, that SW must be compatible with your Uno. It also seems likely to me that an UNO is not a powerful enough chip for your task. I'd suggest using something bigger, say a Mega or a Teensy. You may need to be able to support hardware floating point in a separate chip. Big job, but I like ambitious people. Good luck.
This is a simple auv project where the intended depth is up till a maximum depth of less than 3m deep (swimming pool size).
The purpose of this auv is to maneuver around a mini obstacle course. This is just a simplified version of an auv as this is not for commercial purposes.
Thank you.
1:1:
First thought: "what will the unit actually do autonomously?"
jrdoner asked this also ...
What is the task?
Hi, as mentioned earlier in my reply to jrdoner,
The purpose of this auv is to maneuver around a mini obstacle course. This is just a simplified version of an auv as this is not for commercial purposes.
Thank you.
Cool, sounds fun.
So the only feedback is from those two sensors? That'll be challenging, but both approachable and worthy
But let's me clear, there's nothing else you're going to mention in reply # 14 ?
(please don't take this personally, that kind of thing happens a lot in these forums, your best bet to have the clever people responding is to be clear about goals and intentions from word go and try to keep specification creep to a minimum)
Above the water, I would recommend rangefinders like the Ping or the HC-SR04. Below water, my personal research indicates that laser rangefinders are the ticket. Specifically visible red wavelength as it's supposed to carry better underwater. As for distance I'm not sure how far it's accurate but you can program your controller to ignore the really high numbers as you would only want to avoid an obstacle when one is reasonably near right?
I can see why an accelerometer will be useful for an underwater robot if it's navigating similarly to the way a submarine would, obviously you don't want to list to the side a lot or pitch to far up or down. As far as depth is concerned, considering you've mentioned it will be operating in a pool of known depth, put a fixed rangefinder out the bottom of the vehicle measuring it's distance to the bottom.
Here's where your challenge actually is. Red wavelength laser rangefinder. You're going to need to build at-least one, waterproof laser rangefinder if you want to detect objects under the water. Fear not, for the trail is already being blazed. Check out this post I just found on a DIY drones page DIY Arduino powered laser rangefinder
I have a friend who is considering building something similar to what you have described.. Do you have any pictures of your prototype?
underwater GPS ?
does it have to follow a known path ? or does it have to evaluate it's surroundings and avoid obstacles ?
it is just a pool with a known depth throughout, (fishpond) that is one thing, but a swimming pool has varying slop so the pressure senor might be something to keep AND also pursue other sensors.
it sounds like you need only a few sensors, that is not hard. most of the parts you have already listed, so getting them to work should not be too much trouble.
dave-in-nj:
underwater GPS ?
does it have to follow a known path ? or does it have to evaluate it's surroundings and avoid obstacles ?
it is just a pool with a known depth throughout, (fishpond) that is one thing, but a swimming pool has varying slop so the pressure senor might be something to keep AND also pursue other sensors.
it sounds like you need only a few sensors, that is not hard. most of the parts you have already listed, so getting them to work should not be too much trouble.
Hi, my budget for this final year project is around $100, therefore an underwater GPS is not within budget. And yes, it has to follow a known path - mini obstacle course (I intend to program the ROV to avoid obstacles because it is suppose to maneuver through the obstacle course)
For this project, i intend to only let the AUV travel across the width instead of the length of the swimming pool to maintain a certain depth. Sorry, what do you mean by other sensors? From what i know from my advisors, these 2 sensors should be sufficient.
Could you please kindly advice me on the arduino coding and connection?
And yes, it has to follow a known path - mini obstacle course (I intend to program the ROV to avoid obstacles because it is suppose to maneuver through the obstacle course)
How will the rover detect whether it is on or off the "known path"?
How will the rover detect obstacles and navigate around them?
1:1:
Cool, sounds fun.
So the only feedback is from those two sensors? That'll be challenging, but both approachable and worthy
But let's me clear, there's nothing else you're going to mention in reply # 14 ?
(please don't take this personally, that kind of thing happens a lot in these forums, your best bet to have the clever people responding is to be clear about goals and intentions from word go and try to keep specification creep to a minimum)
Hi , could you be more specific about your question on the feedback?
Alright, noted!
Could you kindly advise on the arduino and connection too?
Thank you!
1:1:
$100 ?
Yes, it is because i am supposed to work on an expansion kit for 8th grade students , where this project is to convert a Seaperch ROV into a simple AUV. As the school will be purchasing the Seaperch and expansion kit for the students, they would not want to exceed their budget. Therefore, I am finding ways to make it simple so that it will not exceed the budget
Yes, it is because i am supposed to work on an expansion kit for 8th grade students , where this project is to convert a Seaperch ROV into a simple AUV. As the school will be purchasing the Seaperch and expansion kit for the students, they would not want to exceed their budget. Therefore, I am finding ways to make it simple so that it will not exceed the budget
You need to provide a link to the Seapearch and expansion kit that will be used. As you don't really know what will be required for the project, a $100 budget may not be realistic. You will need to figure a way to keep the rov at the depth desired. A compass or gyro might be used to maintain the desired heading.
Heelsandsatin:
Hi , could you be more specific about your question on the feedback?
More specific? um, maybe... (?)
I guess it had a rhetorical question element to it - what I was really thinking, is OK, well that's not much feedback, it's certainly something - it may require some sophisticated approaches to optimise it's utility. I'm not sure you have the skills (yet) to work with such a small and potentially noisy set of inputs.
Have you noted the other suggestions for how the 'autonomous' part of the project might be achieved, such as:
- Ping rangefinder, laser rangefinders
- GPS
Chargin suggested it's impossible also...
Within reason the more sensors the better (google 'sensor fusion') - but of course you need the skills to effectively manage all those signals.
Does this make you think, 'yip, yip, cool, time to flex my knowledge!', or more 'hrrrm, what does that mean, might just ignore that...' ?
Hi all,
Sorry for the confusion. To clear all doubts, I am given this ROV Seaperch kit ($134 w/o battery pack) which after it is constructed, come up with an idea to enhance it for the students in their following year, within a budget of around $100 due to the constraints of the school's budget after purchasing the ROV basic kit. The constructed ROV is about 12" long, 7" tall and 7" wide.
Here is the link: http://www.auvsifoundation.org/programsandoutreach/seaperch/new-item
Attached below is a picture of my constructed Seaperch ROV.
I have come up with an idea of enhancing the ROV by converting it into an AUV.
Ultimately, i would like to incorporate the following;
- MPX5050DP IC pressure sensor - to move to specific depth based on the water pressure $ 13.10
- LSM303 LSM303DLHC 3-Axis Electronic Compass Acceleration Module - to direct the auv $10.34
- Arduino Mega - the brain $60
Total: $83.44
The objective of this AUV program is to program the arduino with specific directions to lead the auv through an obstacle course, the auv is not given the freedom to roam and explore the course by itself.
Thank you all for your kind replies.
I an not familiar with the sensors you have listed and what processing they need.
but unless they require some features of the mega. you could use an UNO or a NANO for that matter.
that would free up the costs for an enclosure. figure the smaller the enclosure, the less costs. a simple gasketed enclosure should work in less than 10 feet of water. water sensor for automatic surfacing ?
You're still going to need to give this vehicle a way to know where it is going. The only feasible way I know of to do that underwater is going to be using a laser rangefinder. Ultrasonics aren't going to work well if at all underwater and sonar is going to be well outside your albeit unrealistic budget. Using the fewest sensors possible, you might be able to get away with using only two laser rangefinders, pointing 23.5 degrees off to the left and the right, or a single one mounted on a servo, normally facing forward but with the capability of sweeping to the left and right.
With this second setup, you can easily program it to stop when there is a near obstacle to the front, then sweep left to right and make a determination on which way to go based on those readings. As for the pressure sensor and the gyro/compass, those will give you more accuracy but you might find that you don't need them at all.
My recommendation would be to sort out obstacle detection and avoidance before you worry about depth and heading. This would be fairly easy to sort out with a small radio controller boat with the sensors dropped down into the water.
