Since our last status update we have fully begun with the hardware and software implementations of our project. Over break we all did a lot of research on our various subsystems to ensure that we would be able to jump right back into things when returning. We started by re-evaluating our schedule now that the design report has been submitted and following our individual research, and we set new deadlines that are more realistic and useful for the current state of our project. Over the past couple weeks we began calibrating the webcam for doing 3D hand and object tracking, and we also began implementing basic contact detection and glove tracking. We feel it is important to get at least a rough pass over this part of the CV subsystem implemented early, as once it’s working we can generate some kind of output data to send to the MIDI board. The sensing glove is starting to come together as well. The software to clean up the accelerometer data drift with complementary filters has been written, but unfortunately it has not been tested on the final sensor yet due to issues retrieving data over I2C. The complementary filter has been tested on a cheaper and less accurate accelerometer, however. Work on the RS 485 communication began as well. A test circuit has been set up with two Arduinos, and simplex communication has been achieved. Presently there are issues with the synchronization of half-duplex communication, and debugging these issues has taken most of the past week.
After reevaluating our schedule we are now somewhat behind. The computer vision subsystem is trekking along as scheduled, but the sensing glove and communication protocol are somewhat behind. Unfortunately those two subsystems ran into some rather tricky bugs, but we have planned extra meeting time over the next week to ensure that they are resolved in time to meet our goals for this next week. Additionally, we realized that we have more slack time in our schedule than we initially planned for, so we aren’t concerned about these slight delays.
By the end of the week we hope to have half-duplex communication established between the Due and both of the other subsystems individually. The Jetson will be able to do basic contact detection and output its results over RS 485 to the Due. We also intend to assemble our first prototype of our glove design, and generate dummy output data on the Micro.