The Baskin School of Engineering’s Corporate Sponsored Senior Project Program (CSSPP) fosters relationships with the Northern California business community through student, faculty and corporate collaboration. The corporate sponsor, together with faculty, plans a project on which a team of engineering seniors work to fulfill their capstone requirements, providing the students with the opportunity to gain real-world experience and insight into working in industry. At the same time, the participating companies benefit from the work done by the students.
This year, one of the projects was sponsored by Atollogy, a Software-as-a-Service (SaaS) software company that offers proprietary algorithms and capabilities for managing operations by integrating the physical world with artificial intelligence. Students An Tran, David Munoz, Cesar Neri, Kevin Ajili, Eric Su, and Arindam Sarma worked with Atollogy on research that analyzes data from production lines.
The team spent their first CSSPP quarter working to help an Atollogy customer understand the flow and activity of cement trucks in their yard. Using advanced image-recognition algorithms, the students researched and tested methods of license plate recognition, providing Atollogy insight into the most effective solutions to their problem.
The team then worked with Atollogy on the Greens Only project, the goal of which was to find an efficient and effective way to identify foreign objects in a produce production line. The team focused on buiding an automated detection system specifically for spinach. Because spinach is always a uniform green color, foreign objects, which will almost always be some color other than green, are easier to identify.
Ultimately, the solution depended on a combination of hardware and software solutions with computer vision algorithms.
To replicate conditions in the real facility, the students had to create their own data in order to test the software. “We wanted to research different environmental conditions, foreign objects and contaminants,” said Cesar Neri. “To do this, we needed to create our own ‘contaminated spinach’ images.”
The team explored multiple methods for the foreign object detection before coming up with a solution: the creation of a pipeline to process images from a video feed of a spinach processing line. The team took the video footage, split into individual frames containing about 1,000,000 pixels each, and masked the static background. Then, they filtered out all of the green pixels in each frame. At this point, only non-background, non-green pixels remained. Remaining regions of contiguous color above a certain size were marked and saved onto the computer with the filename serving as a timestamp for when the potential foreign object was detected.
Although challenging, working on the project was rewarding because the students were able to experience working closely with a Silicon Valley company, and were able to see the results of their hard work implemented in technology used by Atollogy. The project also provided the opportunity for networking, and three members of the research team will intern with Atollogy this summer.
“I thought it was a really cool experience because it was my first time doing practical work for a real company. Seeing my work, time and effort going into something that’s not just a grade felt really good. It was probably the most important experience during my studies,” Kevin said.
Atollogy was pleased with the accomplishments of the students. “The students have been wonderful to work with," says Anthony Tarantino, founding member and senior advisor to Atollogy. "They were not afraid to take on what is a very complex problem. They've been enthusiastic, professional and well-organized throughout the project."
Tarantino, who graduated from UC Santa Cruz in 1971, said, "It's great to come back and bring a project like this to fruition.”
Atollogy enables manufacturing companies to enter the modern world of technology without the need to endure the pain, expense and process changes that are the hallmarks of classic enterprise systems. We offer proprietary algorithms and capabilities to revolutionize how physical operations are managed by integrating the physical world with artificial intelligence. Our goal is to let our customers leverage enterprise-class computing & analytics with the ease & simplicity of consumer technologies to improve their operations. Atollogy is privately held startup with offices in Santa Clara, California. https://www.atollogy.com