This is a fairly amazing take on auto-rendered custom videos than has nothing to do with digital signage, but should certainly get some technical and creative people thinking about the possibilities for mass creation of creative spots.
Adweek has a post up about how Adidas created more than 26,000 customized highlights videos for every runner who crossed the finish line earlier this week at the Boston Marathon, cranking them all out in 24 hours.
“Our idea was to harness the power of 30,000 runners—generating data from the very race bibs that sport the Adidas logo—and transform the entire race into a creation engine. The result is an epic personal highlight film for every single runner who becomes part of the legend of this incredible event,” Drew Ungvarsky, CEO and executive creative director of Grow, told Adweek.
Grow had a crew of 18 people with eight cameras along the course, four getting b-roll while the other four captured personalized footage. In all, they had 27 hours of footage.
Certain checkpoints triggered the RFID chip on each runner’s bib. The RFID chip not only tracked their pace and time, but also helped Grow locate runners in the captured video footage. The agency used custom software to grab video footage that fell in line with the time that the runner crossed a certain checkpoint. They could the programmatically zero in on the few seconds that the correct runner was in the frame and include that footage in the personalized video.
The final videos have the runner’s name, their times at different checkpoints in the race, as well as a personal clip of the runner taken at some point along the course.
Not sure how the videos were rendered, but would assume there is a base video template with the b-roll, and placeholders for the runner name and so on, the interval and finish times, and the personal footage, that all mashed up by software application and probably a cloud-based rendering engine.
Extend this sort of thing out and you could imagine who, with permissions in place, a digital OOH campaign could be booked that runs versions of these videos in local markets, using runner entry information to identify a runner or runners from that market.