Honors Link

May 9, 2012   Over past year, I have experienced a remarkable turning point in my academic career. For the first time, I was able to apply all that I have learned over the years to two projects, my Honors College Thesis and my Engineering Senior Design. Both of these yearlong projects required intense focus and determination. In addition, my Honors College Thesis tied directly into my Engineering Senior Design project.

My thesis consisted of finding a low cost and accurate method of locating one’s position. I began my research by looking into radio transceivers. I had planned to combine two methods known as trilateration & triangulation with three transceivers to accurately locate your position. After working out the math however, I realized I would not be able use any old transceiver. It turns out that standard 2.4GHz transceivers do not have the right bandwidth required. In addition, the micro-controller I would need to use would need picosecond clock cycles. Although, it would be possible to design and build a Field Programmable Gate Array (FPGA) in place of the micro-controller, the time required for this is beyond the scope of an undergraduate thesis.

The next step I took was to research Ultra-Wide Band localization technique. I still had planned to combine the two methods of trilateration & triangulation however, this time with three UWB receivers instead. UWB receivers have the necessary bandwidth required. Plus, if I were to use UWB receivers, a high-end micro-controller would have given me the speeds I needed.

The best part about using UWB is it can be used to triangulate a position down to only a few centimeters! With this in mind, I worked out all the math, created MatLab simulations, and began researching UWB receivers. As it turns out, UWB receivers are most certainly not low cost. In addition, the range that most of these receivers had was less than 10 meters.

After discussing these issues with my mentor (Professor Huaping Liu), he pointed me towards a company called Time Domain. It turns out TimeDomain.com has exactly what I was looking for. UWB transceivers capable of communicating with other UWB transceivers over distances greater than 100 meters and finding the relative position down to a few centimeters. Eureka! The transceivers are called P400-RCM UWB Transceivers. Unfortunately, for what I needed, it was going to cost $20,000 just for the transceivers.

Without giving up, it turns out the only method for relative localization that is available for relatively low cost and fairly accurate position is a Global Positioning System (GPS). However, low cost GPS modules are only capable of approximately 3-10 meter radius positional accuracy. For my Senior Design project, this was not enough.

In order for me to improve the accuracy of a standard GPS module, I opened my Artificial Intelligence textbook and began to implement a machinelearning algorithm. Using a Gaussian distribution and a utility function to keep

score, I was able to improve the accuracy of a standard GPS module by about 75%! My method was to place three GPS modules (of the same type) at the corners of an equilateral triangle with side lengths of one meter and implement my machine-learning algorithm. After months of coding, it worked like a charm. So how does my thesis, Triple GPS Localization (TGPSL), fit in with my Engineering Senior Design project? My Senior Design project was to create a radically new way vehicles operate. In other words, creating a smart-car. My Senior Design project was to design and build a Platform for Autonomous Multi- Vehicle Communication and Coordination (PAM-VCC). PAM-VCC is a platform that enables other cars equipped with PAM-VCC to communicate their velocity, acceleration, direction, weight, braking distance, and location to each other. Essentially, PAM-VCC created a network of cars. This information is then displayed on a Graphical User Interface (GUI) for the driver to see.

Imagine driving around a mountain pass and knowing if there is a car around the corner. Imagine driving through dense fog or a blizzard and being able to clearly identify all surround vehicles within a 500-meter radius. In addition, what if your car was able to identify a car accident ahead and tell you to slow down or take a detour? Well, I will go one step further, imagine your car begin able to predict if a collision was going to occur and stop it from happening. PAM-VCC is capable of taking control of your car when the artificial intelligence agent inside detects that a collision is about to occur. However, for all this to happen, there is one very important detail, location, location, location! It is vital, that the PAM-VCC equipped vehicles know their exact location within a

few centimeters. However, as an undergraduate student with limited time and funding, I settled for my 75% increase on the GPS modules. Utilizing my Triple GPS Localization technique, I was able to locate the position of the vehicles within one meter 95% of the time and within half a meter 50% of the time. Although, this is not accurate enough to fully implement my Senior Design project, it was enough to build a concept prototype and that is exactly what I did.

There is still much work to be done for both my thesis and the senior design project. If another person would like to continue my work, I left a few ideas at the end of my thesis. In addition, my Senior Design project will remain on the Engineering Beaver Source webpage for the years to come. Anyone is welcome to improve on our current concepts or extrapolate an entirely new method for PAM-VCC.

The HC Experience scholarship greatly contributed to the success of both my Thesis and my Senior Design project. Without the funds, I would not have been able to complete either project. The Experience scholarship helped my pay for everything from nuts, bolts, wire, and solder to the wireless communication, accelerometer, vehicle platforms, and the GPS modules themselves.

These two projects truly combined all the knowledge from all my classes taken here at Oregon State University. Since my passion for engineering focuses on artificial intelligences, robotics, and control, both my thesis and the senior design project played a significant role in enhancing my academic success as well as my future career plans.