Rohit, Berkeley Deep Drive, Week 3

My main project is working on creating an accurate vector map of the entire Richmond Field Station by using the relatively inaccurate GPS coordinate from google earth/maps. I'm working on this project with Devin who is a local rising senior and Heiru who is a graduate student originally from Taiwan but attends the University of South Carolina.

A vector map is the defined area that a vehicle knows it is allowed to operate in. The benefit of having a vector map is that an autonomous vehicle can still function in areas where there may not be GPS connection due to interference from buildings or trees. The vector map is also considered to be more accurate than GPS, but the trade off is its high creation cost and time requirement, as there is no public software that can accurately make a vector map from raw collected data. The data that will be used to construct the vector map was collected from the test vehicle as it was driven around the facility. Here is a picture of the raw data in a compressed 2D image. The image shows one of the main issues we ran into, specifically the autoware tool that compresses the raw 3D data into a 2D image stacks the points together creating significant inaccuracies. Therefore it was very difficult to even attempt to make a accurate vector map using this method.

The attempted vector map is drawn in purple and the actual map of the facility is drawn in green via an image from google maps. This highlights the discrepancy between the compressed 2D point cloud image and the actual image from google maps.

Comments

Evan Bradley, Week 3 at the Missouri Orthopaedic Institute

As mentioned in my previous blog, I have been awaiting ligament, menisci, and cartilage tissue from a canine or human knee joint for (interleukin) IL-1B tissue culture. IL-1B is an inflammatory cytokine that has been proven to increase rates of tissue degeneration and osteoarthritis development in the Thompson Lab. Dr. Stoker wants me to experiment with different types of knee tissues in a co-culture with varying levels of this cytokine to determine its effects on the entire knee joint. This co-culture uses an insert permeable to the media to separate the two tissue samples from physical contact, while allowing them to share the same media. This creates an extremely accurate model for knee tissues in their native environment due to their exposure to the same synovial fluid in the joint. This model would then be treated with the IL-1B and cultured for 21 days. During these 21 days, the media would be collected every three days for biomarker evaluation at the end of the stu...

Kylie Heering, Week 2 at the Goldstein Lab

We started off our week with a congratulatory acai bowl trip to celebrate Preston’s acceptance into a training grant program. Acai bowls in California top Playa Bowls (no question about it). From what I can tell, its a pretty huge honor to be recognized by this grant, but he’s really humble about it. On Monday, Preston and I decided that testing antibodies that have never been tested on prostate epithelial cells before would be a good objective for my first Western blot on my own. We needed to probe for ASCT2, a glutamine transporter, and GLS in order to determine if their corresponding antibodies are functional. Antibodies are crucial for Western blots because they bind to the protein of interest (POI), allowing for us to qualify its expression after imaging. As such, Preston wanted to make sure they worked by probing for ASCT2 and GLS on three different cell lines. Cell lines are commercially purchased human cells that have been immortalized (modified to grow indefinitely) by telome...

Daniel Cheng, In My Own Room

I had no idea that Pennsylvania is this wide. Within the first hour, my train had reached Philly. But to Pittsburgh, it took another seven. Even before I stepped foot into the Search-Based Planning Lab, I was waylaid by some anxious news. The PhD student assigned to be my mentor, Dhruv, texted me that he, Dr. Likhachev, and most of the lab would be out of town for the entire week. So that was that. Fortunately, it was my first week, the week to be spent learning new material, and Dhruv provided me with plenty to digest. I already had ROS (Robot Operating System) installed, so I looked towards the tutorials that ROS provided. I copied commands into my Linux laptop's terminal to run ROS features. I learned the basic structure of ROS: packages, services and clients, publishers and subscribers, messages, nodes, and topics (which nodes communicate messages over). There was one simple yet interesting program I came across in the tutorials called turtlesim, for which using only 2 comm...

Peddie EXP Summer 2018

Search This Blog