Embedded Files
progress
progress
Our main accomplishments this previous week were as follows:
- Achieved control of the robot using Python
- Got approval for the general power system by Brian Jensen
- Successfully powered arm using robust 110V to 240V transformer system
- Created basic control GUI for testing arm control
We have made solid progress on automating the arm from the computer. It turns out that you can effect basic control over the robot using roughly 6 lines of Python. You can see this in the video on the left below.
We also spent time starting the design and planning of our safety systems. We replaced a key'd receptacle that came with the robot (and was wired poorly) with a reliable Estop system. When activated, this system will immediately (see the video on the right below) stop robot motion by removing power to the arm, which will stop the motors and engage the motor brakes. Attaching the new Estop button required a small amount of CAD and 3D printing (seen in the image to the right).
Adapter coupler for Estop switch.
We have iterated the design for the power system for our robot, and we feel that we have arrived with a system that is significantly more robust, and more importantly, safe, than our original design. We found a 110V to 240V transformer (P/N P-8689) that is rated for 1000VA, or 4.35A at 230VAC, in the Applied Robotics lab. Originally, we were concerned because this was under the rated max load of the arm (roughly 1.5kW), but upon reflection we have determined that this transformer will very likely suffice for all of our needs, with a wide engineering safety factor. This is largely based upon the valid assumption that we will be using effectively no load on the scale that the arm was designed to operate during.
We will also monitor power usage throughout the development of the robot to ensure we do not get close to the limit. If we do, there are alternatives that are easily within our budget. Current experiments show that the arm only draws ~125VA at idle and small amounts of motion (see image to the right).
Power used at idle/slow motion.
Initial voltage testing showing transformer output voltage under a light load.
Transformer box wiring using industrial DIN gear.
We have written a basic GUI that replicates some of the basic functions that the DENSO physical pendant provides (that we do not have access to on a regular basis). This allows to move to specific points now, and will add additional functions as we write the code for them.
This software also lets us test the software library that we are slowly building up for the arm.
Challenges
Challenges
We've had a few challenges this week, mainly surrounding our power system. We have started the research to make sure all of our wiring and system setup is compliant with applicable standards, but that will take some amount of time to complete. It will be done for sure by the seeding competitions, and hopefully some time before that.
Additionally, there is a small utility provided by DENSO (visible to the right) that allows us to enable arm motors, and we are working on finding a way to remove this requirement, as we'd ideally like to run Ubuntu on this system and use ROS, and the utility is windows only.
PLAN for next week
PLAN for next week
During the next week, we plan to accomplish the following:
- Make a functioning proof of concept of our latest actuator design
- Enable arbitrary control of the arm
- Finalize our arm mounting design to a robot chassis
- Design a sensor to determine if a ball is caught
- Mount the proof of concept onto our arm, integrate it with a sensor
Google Sites
Report abuse