What do Robots, Stem Cells and Photography have in common?

Ever wondered what it would take to make a robot that can avoid obstacles? How can computers make you a better photographer? What can technology do to improve stem cell research? These are just some of the projects going on at Intel Labs Pittsburgh.

I recently attended the lab open house. This annual event is an opportunity to experience the full range of research activities and collaborations, via posters, demonstrations, and direct conversation with the researchers and students involved. On display were the latest work with Carnegie Mellon, University of Pittsburgh, and many others. Research projects covered topics from personal mobile robots, computer-assisted medicine, parallel machine learning, power efficient server design, cloud computing on big data, cache-savvy algorithms, using solid-state disks for databases, multicore system design, and optical interconnects for datacenters.

Collaborative Robotics

Researchers from Intel and Carnegie Mellon University want to make robots common in people’s everyday lives by increasing their usefulness and acceptance. They acknowledge the current limitations of robot autonomy and explore methods for effective human-robot collaboration. There goal is simple – robots helping people & people helping robots.

To accomplish the goal of operating cooperatively with people, robots need to overcome a new set of challenges, in addition to the standard difficulties of perception, planning, navigation & manipulation. High among these are ‘people skills’ – detecting, recognizing and understanding the actions of humans in the environment. We are investigating opportunities for collaboration when the human and robot are near each other (local collaboration) and when they are far apart (remote collaboration). We are developing a rich and modular set of robot skills common to multiple applications. Such skills enable robots to provide new compelling services.

Better Photography thru computation

We all take lots of photos, but they don’t always turn out that great. Wouldn’t it be nice if there was a way to have your computer help you become a better photographer? This project using machine learning, spatial recompostion, traditional photographic aesthetics (Rule of 3rds and Fibonacci’s Rule) combined with minimal user interaction to improve photos by 73%.

Stem Cell Tracking using Computer Vision

In the past, biologists have had to manually track stem cell proliferation but now by using computer vision algorithms, they are able to achieve the same accuracy levels using computers. This project was also displayed on a remote screen using the Open Cirrus cloud computing testbed. Being able to check on a project remotely will remove time constraints and enable biologists from around the world to utilize these tools for their experiments.

Comments are closed.