A Challenge to Universities: Innovation Isn’t Always About What’s New
July 01, 2016
Guest Blogger: Graham Fletcher, Manager, Cubic Innovation Centre, Cubic Transportation Systems (CTS) U.K.
One of the best aspects of working in innovation is spending time at university research laboratories. It’s at places like the University of California, San Diego’s (UC San Diego) Center for Visual Computing that the next generation of basic research is being conducted.
Just over a year ago UC San Diego opened its new venture, the Center for Visual Computing, which Cubic sponsored as a long-time UC San Diego supporter. One of their first tasks was to address two related problems: computer vision, which is the act of getting information into a computer from visual sources; and computer graphics, which applies to obtaining information from a computer into a visual form. A year on, we recently attended the first annual review of their work – and we were amazed. And, admittedly, it didn’t hurt that part of our meeting with the team included a walk on the local beach – a rarefied experience for those of us from rainy London.
During the visit, we discovered some wonderful ideas that will be making their way into CTS’ technology trials this year. But, it’s interesting and a little frustrating to know what isn’t there – yet. Research by its very nature concentrates on doing more with the best and newest capabilities. However, one of our challenges in public transport is that infrastructure tends to change very slowly. There are 1.85 million closed-circuit TV cameras in the U.K. alone. These devices are unlikely to be upgraded to become 4K ultra high-definition cameras and nobody is going to provide professional photo studio lighting in the stations. Our challenge is not only to do more with the best and newest, but also to do better with our installed technology base.
This is an invitation to academia around the world: I don’t only want an even better 3-D facial model from a single camera; I’d also like you to get me one from an old black-and-white CCTV camera in bad lighting. Now that’s a challenge!
You might be asking: computer vision and transport together? Why? I admit these don’t seem like obvious bedfellows. It certainly hasn’t been a focus for us at Cubic and it’s not part of a core transit system. Indeed, it’s always been a peripheral function in the security space, more often used by police for monitoring the crowd than delivering transport operations. So, are computer vision and computer graphics part of the core delivery of a future transport system? You’d better believe they are.
Visual computing could be the future of all sorts of directed and personalized services, fare collection, and even provide the data needed to plan the build of future stations. We need greater understanding of passengers and their behaviors and choices, so we can better align transport services. If a computer could “see,” what might it add? It could tell you how busy your station is and give you an idea of who uses your station? Visual computing can lend insights into why people are traveling (differentiating between commuters and leisure travelers based on observable behavior) and could also provide invaluable services like identifying children and providing navigation to the lost. Visual computing might recognize a person as somebody who’s paid their fare or as someone who hasn’t, or even as someone who habitually never does.
Station staff are trained to perform many surveillance functions. They learn to recognize the signs of someone thinking about jumping from a platform or how to differentiate between a lost suitcase by a regular traveler and a planted bomb. But staff can’t be everywhere and see everything. Visual computing researchers are trying to solve the same problems. I look forward to the day when station staff members are alerted to a distressed person on a platform, while a friendly “Hello, can we help?” still makes a difference.
We have a fantastic relationship with UC San Diego, and we are eager to build partnerships with other research institutions. So I’d like to extend an offer to all. If you think you have a technological solution that can further our requirements in visual computing, then we’d like to talk.
Southern California sand is not compulsory.
Tags: Cubic Transportation Systems
, Cubic Corporation
, UC San Diego
, visual computing