The following are the notes from the meetings this year to date:

5/1/2003:
The new Mark III bot is being fitted for its new hardware attachments by Douglas. Sensors currently include a Devantech servo-mounted sonar sensor, three Sharp analog output IR sensors, three optical mouse-based sensors, two shaft encoders, and the 802.11b wireless LAN PCMCIA card that will be used to detect signal strength, but also for data communications when a strong signal is present.

The majority of sensor/motor interfacing will take place on the PIC16F877 that is included in the Mark III kit. Roman is in charge of that portion of the implementation.

Jacob is working on stochastic models for robot planning and navigation, and is generally overseeing the project.

Zac is using an ADAPT812DX board from Technological Arts to interface with the PIC and 802.11b Prism card, and also to store and transmit the data. The board is based on a Motorola HC12 microcontroller, and it has 512K of external SRAM and 512K of external flash. Mike Meeuwsen is assisting with some of the C programming on the 802.11 interface, as he has experience in that area.

3/16/2003:
Very interesting things have been going on lately. We came out with a design plan and a list of parts. After we completed all our research we realized that all of our materials and equipment would cost over $2000. Since there was no forseeable way we could get this amount of money in the immediate future, we decided that we needed a scaled-down prototype.



We are currently using the base system from the MarkIII sumo kit issued by the Portland Area Robotics Society (PARTS). We are modifying it to suit some of the specifications of the ANA project. Most notably we are going to test our optical navigation design using Agilent HDNS2000 chips. We will also include minimal sensors such as IR and a single rotating Devantech sonar sensor.

2/21/2003:
Currently looking to acquire parts and donations. The general plan is all drawn up but we need the equipment to put it together. A list of needed parts will be posted soon.

1/8/2003:
There are two extra mice that contain the HDNS-2000 chip. Zac will be working on interfacing one with a microcontroller. Douglas is going to begin work on the working base.

Winter Break:
Jacob prototypes an optical sensor-equipped cart:



"This is a picture of the optic sensor attached to a cart. I assembled this with the help of Douglas and a couple other people who emailed me with some practical advice.

We discarded the brass tubes because they were impossible to get the aluminum rods through. This was because the end became tapered after it was cut, and no amount of filing and deburring got it back to a perfect shape.

In order for the optic sensor to work properly, it had to be 2.4mm +- 0.2 mm above the surface of the ground. This was especially tricky to achieve, but we were able to adjust the base clearance by putting an AOL cd between the axle and the base as can be seen in the picture.

After we plugged it into a computer, it moved up and down just like a normal mouse would. The resultant clearance from the lens to surface ended up being 2.1mm. It's a little closer than we wanted, but the side effects aren't immediately apparent."

12/3/2002
In small quantities, optical mice appear to be the cheapest source of parts. Jacob will be ordering some over break.

11/26/2002
The ADNS2051 from Agilent seems to have all the functionality we could want.
(It's the low-cost chip too!)
More optical navigation @ Agilent
More notes soon. (There's an alarming shortage of scanners at Hovland).

11/19/2002
A bit of change of pace as we decide to get our hands dirty.
Summary

11/12/2002
List of things to further break down
Sensors List
Sensors