Jack and Dale have submitted their undergraduate project proposals, which were accepted, and will convert the to either HTML or PDF format to go on web.
Everyone is busy. Chris has been working on the H-bridge interface circuitry to the computer and providing sufficient noise reduction/compensations. He is waiting for parts to complete the prototype and for a go-ahead to put a thruster in the tank. He has also been working on further literature search and for any papers on AUVs published on the web. Jack has been reviewing the Sony manual and the existing device driver for the Sony EVI-D30. Dale has been collecting information on sensors to use for proprioceptive sensing. Alex has been whipping the building contractors and will look into some sensors he knows about in Japan. David has finished up with schedule and budget and has been collecting information about the computing system.
The paper "Towards Terrain Aided Navigation of a Subsea Vehicle" by Newman and Durrant-Whyte provides a good overview of the sensors installed in the AVU Oberon. Observations: A spaceball may be a good control input device for our sub. Accelerometers and gyros provide good high frequency information but drift quite rapidly. The gyro and accelerometer are digitized at 14 bits (wouldn't 10 bits be sufficient?) A combination of these with magnetic compass, inclinometers, and depth sensor may provide a robust overall package and Oberon seems to have had some success with this approach although drift is still proportional to sqrt(time). High-resolution sonar will not be our initial focus and interestingly it seems to have problems with diffuse features (like sea weed and barnacles) as well as side-lobe reflections. The system model (equation 1) seems a reasonable place to start but information on how it is formulated is required. [Next week we'll look at another paper by Durrant-Whyte on this topic.] It is noted that thruster turbulance causes problems with underwater imaging--this could influence where we place cameras.
The design of the computing system for our AUV [name yet to be determined] is described in design notes. The design is incomplete but it seems that there are sever options for each component that meet our requirements.