Roger: 13th July 2007

1) Hardware

a) Robotic Arm
i) We had initially wanted to increase the speed of the arm by (A) using a gear box & (B) changing the motor to a lower voltage motor. However, option (A) failed to work as the increase in speed led to a reduction in torque which prevented the lead screw to rotate. Also, option (B) did not work as planned due to failure in the motor to rotate.
ii) Finally, we decided to increase the speed by changing the lead screw system with a pulley system. The pulley system can increase the speed considerably. This is given by v = r w, where v is the linear speed we are trying to achieve, r is the radius of the pulley and w the angular speed of the motor. To reach the speed of about 10 - 15 cm per sec, and with the current angular speed of 272 rpm, we calculated the radius to be around 5 mm. Also, the amount of torque required to drive the pulley would be equivalent to the frictional force from the pulley. This can be reduced by using bearings and pulley. This has been implemented and is working properly.
iii) Further work is to be done on the z arm. This is because the z arm cannot be replaced to the pulley system. We will be using the gearbox but using a smaller up ratio. This was tested and is working. We have approached the lab officer in the ECE workshop to help us make a coupling to connect the gearbox to the lead screw.

b) Integration of the arm with the human avatar and virtual terrain
i) Work had been done previously to integrate the arm with the human avatar and virtual terrain. To work on integration with human avatar first to match the speed of the human avatar with the arm.

c) Backup list
i) Most of the components for the backup list have arrived and will proceed to solder on the extra slaves and masters for testing after the full system is up and running.

d) Camera tracking
i) White tracking is now complete. Another application based on average color of the background has also been done. This application is much suited to low lighting environment since it tries to distinguish objects from the background. This should work with nonwhite hamsters too. But this has to be tested.
ii) Completed histogram based tracking application. Histogram based application may be difficult to use since we do not know the histogram of the hamsters brought in USA.
iii) Removed the counter display from the main window and added a frame rate display. The frame rate is around 15. The white color based tracking program is smooth and does not have any delay problems.
iv) When the side panels are fixed to the structure, lighting around the corners is very low. Therefore the tracking suffers. Went to Simlim and bought a “down light” to test with the system. It seems the down light is not sufficient and more powerful light may be required for white tracking.
v) Most of the week is spent on getting a proper lighting system ready. Some of the lights bought from Simlim give reddish light, which cannot be used. Finally 12 DC LED light strips were selected. They give a bright white light. LED strips were fixed to the structure using metal plates similar to lighting in a stadium.

2) Animal training

a) To satisfy the requirements for clearance from Institutional Animal Care and Use Committee (IACUC) and Office of Safety, Health and Environment (OSHE) for using hamsters for the project, Roger and Janyn will be attending a 2-day training certification course on 16-17 August.

3) Hamster training

a) Chasing attractor in structure with moving actuators: Meena, J and E gave satisfactory results (average results were above score 4) in chasing and tunneling inside the structure. Food was given occasionally inside the tunnel as their reward and encouragement.

b) Study 2: The results were not that satisfactory. The hamsters went into each room equal times. They liked tunneling through the tunnel divider and the attractor, and occasionally run on the wheel inside the room. Improvements to be made: (A)to cover the other room opening once they choose which room to enter (B) to forgo the divider, put the wheel, food and the moving attractor inside the big room to (C) do not use tunnel for hamster to choose the rooms, instead use black or white door indicator.

c) As we will be using the black boards inside the structure, we will need two doors and a tunnel in between for the hamsters to get from the cage to the structure. The structure door is to prevent the hamster from hiding inside the tunnel when the gameplay starts.

Summary of the meeting on 10th July as summarised by Janyn

Updated timeline as given:

timeline1.jpg
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-Share Alike 2.5 License.