Some of the robots from
Some of the robots from RoboCup at Home might be a good inspiration for things to do with your project. After all, what good is a design without requirements?
Good Point IG
I watched the video … look at all the Kinect sensors strapped to heads !
You have a good point regarding requirements. The software requirements I have for MRL are quite extensive… One of the most important is that it will be able to work with multiple platforms and actuators (arms being the current topic) with very minimal changes (if any)
The arm’s mechanical/physical/sensor requirements should model the functional capabilities of an arm. (I’m interested in function - not interested in making it look like a person, Markus may diverge with me on this, I’ll wait for his input) I think human functional requirements for an arm would be multi-jointed (could be super jointed snake/elephant trunk like), capable of easily lifting and placing a 10lbs object from ground level to above counter top level ~50", Grippers would need to be able to grasp a large variety of objects (fingers, suction cups, coffe filled balloons, other?)
I guess my point is that
I guess my point is that requirements like that are OK, but lack a certain level of connection to real world tasks. Challenges on LMR and contests like RoboCup get people focused on specific tasks, which can be a great help while designing and implementing.
As things develop for the MAAHR, consider looking around the Internet for challenges that it might be able to particiate in. Even if it doesn’t get entered, if you adopt a set of tasks to achieve, it will provide a lot of focus for the team.
Or take the requirements you’ve already stated, and design a set of experiments/tests around them. Can MAAHR get you a cup of coffee? Can it pick up that egg without breaking it? Shake hands firmly, but gently? You get the idea.
Out of context
I only responded about arm requirement in that way because that was the initial inquiry from JoeBTheKing. And I responded with valid generalized design principles. At the moment the current focus is still with the brain. It’s the only part which has materialized. It will take some time to get it prepared to a capable platform level. This is the next “real” milestone.
The next milestone will be testing some of the MRL interface connections to several different micro-controllers.
The subsequent milestone is really getting a base and constructing the necessary interfacing. There is a new differential drive service in MRL - which probably needs more refinement.
The next milestone would just be simple obstacle avoidance.
Next - mapping, slam, object detection - The goal at this point would be to find a electrical outlet and recharge. The find and recharge would be the first serious challenge.
I’m working on the brain still, but need a base ASAP. Markus said he was going to make one, but has recently been pre-occupied. I have put a rudimentary base together, but would welcome any input.
I agree wholeheartedly regarding specific challenges & milestones. At the moment we don’t need an arm, but we need a base !
For the base, I will be using parts which are accessible too me now. This may not be the “optimum” base, but a physical vehicle will help me develop and refine the software so that it will work in a generalized manner (hopefully on multiple platforms). This will keep expenses down and at the moment be the quickest implementation. It can be reviewed, extended, or replaced, but purchasing or shipping new equipment would need to be approved by Markus.
My biggest “value add” is software. In fact I would be more than happy to ship the brain off to someone who can provide more than my capabilities regarding mechanics/electronics/etc… as long as they have a router which will allow me to get back into the brain.
BASE
I’ve ordered the parts for the base today. The base will be 3 wheeled, capable to drive around 10 30kg, complete with wheel encoders and motor driver (I2C).
Next step will be to built a robot arm with gripper, able to lift around 0.5kg. The gripper will be equipped with force sensors (I have the sensors already).
**I’m building a new arm… **
I’m building a new versoin of my 6 axis robot arm, that would be just about right for this project.
Hi Toby,Cool. Do you have
Hi Toby,
Cool. Do you have already some specs?
TROBOT arm specs
The new one should be about 12 to 16 inches tall, with a reach of about 18 to 22 inches. I’d like the payload capacity to be about 2lbs. It’s still in concept and sketching stage, but will be complete by the end of this year (my Kickstarter project is gonna be funded on 8/7 ;-).
You might know that the
You might know that the MAAHR project is sponsored by me.
I saw a cable for the
I saw a cable for the PSU:
It says it’s an extension but should work.
Thanks thernxt123
Unfortunately it needs to be Male to Male…
Maybe that’s why they don’t include it ? Dunno, anyway I managed to slice two male cables off of some abandoned/broken ATX PSUs … will be splicing them tomorrow…
Arms…
The gripper on the end of the arms would just be for testing correct? For the MAAHR project we will probally just use your hands, right?
What Amperage motor driver
What Amperage motor driver will this need? I would be delighted to offer free PCB’s for MAAHR, as well as the free motor drivers.
Hi Krumlink,Thanks a lot for
Hi Krumlink,
Thanks a lot for your offer. The prop system I ordered has already a motor driver, but we will need a lot of other boards for actuators, sensors, micro controlles. I’ll sure came back to your offer.
My hands?:DThe robot hand
My hands?
The robot hand I’ve built is not capable to grab especially smaller things. I will go probably for a 3 finger hand like that: https://www.robotshop.com/world/kinova-technology-jaco-robot-manipulator-academic-edition-2.html
A few thoughts about the brain case
Wood is OK beeing lightweight but I recommend some modifications:
- the motherboard should be down and the HDD and PSU up;
- drill some big holes or cut some openings in all the decks to ensure enough flow for generated heat;
- put rubber spacers on the screws that fix the motherboard (in both sides) for external vibration and shock damping. Also ensure at least 5mm clearance under the motherboard for air flow;
- put rubber spacers for the screws that fix the HDD and ensure that you have at least 5mm clearance under the HDD for air flow. In final version the HDD should be fixed with rubber bands not screws to lower the vibrations and the noise;
- put spacers under the PSU to ensure at least 10mm clearance for air flow, the PSU will probably generate the most heat.
If in the final version the brain case will be a closed box, I would use two 120 mm quiet fans (one in the bottom and one in the top).
All are Bases belong to Us !
I realized (duh) that I have a base,
Although it is a bit too big for the inside of my house, it led me to think of another great challenge/milestone for MAAHR. Willowgarage with it's $400K PR2 successfully got a beer from the fridge. But what if you are out of beer?
Out of Beer Challenge - The Next Step
The challenge would be to navigate from my house to the Fred Meyer Grocery store - using a variety of short range obstacle detection (vision primarily but possibly augmented with ultrasonic and/or other methods) - and long range route planning (google maps, GPS). Arrive at the destination, enter the store, find the appropriate beer (I like black butte), pay (preferably the automated check out line - cause that would be a complete hoot) and return with said prize.
I would like the opportunity of the MAAHR brain & software to be modular and well designed enough to switch between Markus' base and this one. Markus' base could be used to train inside my house and with the minimal amount of configuration change - the brain and software should be compatible with the larger outdoors base.
This base currently needs motor controllers (KRUMLINK!) and 2 new Marine batteries. It's a heavy persons wheel chair and it has a very heavy duty frame and packs a whallop. I'll post specs soon, but I'm thinking 30+ Amps more like in the 60 range. Whatever motor controller is used, it will need a fuse or preferably a reset. Also, getting this large means safety provisions.
I was accepted into the DARPA Grand Challenge back in 2004. We failed to fund a vehicle, however, I would recommend that we follow the same guidelines. There was a requirement that the vehicles have a kill switch which is very visible and readily accessible. The kill switch kills all power to the vehicle, and will not reset without manual intervention. There also needs to be a remote kill switch - which again, takes manual intervention to reset. A tracker (vehicle/person) will be required to follow MAAHR on its beer run with their hand on the remote kill switch.
I'll post more details design, challenge details, motor specs and other parts as I get them.
I can design the board and
I can design the board and send it to the project centered around the HIP 4082 driver, It can handle 1KW of power.
This should be posted as a real challenge
This should be posted as a real challenge, with a really cool prize. Although I think that performing this feat is a reward in itself (the beer is a bonus…)
Thanks for the suggestions cr0c0
- Agreed in that the MB needs ventilation and the HDD should be mounted in a way which minimized gyroscopic effects
- the sides are open, so there (at the moment) is considerable ventilation
- skipping this one until i find some appropriate material - (haven’t found it yet - but might get some rubber gaskets)
- I was surprised how much noise was created, but it makes sense in retrospect because it was firmly attached to a sound board - I put a non-conductive foam and it seems to have exponentially reduced the noise from the harddrive. I will be looking into the processes which do writes/reads and shutting down all unnecessary daemons etc.
- When I get the rubber gaskets I’ll do this, right now as it stands the PSU and motherboard are very cool. The HD is the warmest of the bunch
- Depending on how it is attached to the base, I’ll put some sort of fan on it…