The initial "robot" created by this project was described like this:
"Our first swimmer study successfully demonstrated that the bots, modeled after sperm cells, could in fact swim," Saif said. "That generation of singled-tailed bots utilized cardiac tissue that beats on its own, but they could not sense the environment or make any decisions."
A fair summary would be that the team stuck some cardiac cells on some kind of structure with the result that it moved in a way that mimics swimming to an outside observer, but of course was simply pointless movement because there was no way to control the constructs.
There also is no mention of longevity, usability for any purpose other than "look, it sorta swims", etc. For all we know, these "robots" might only "live" for a few minutes.
The addition of optic stem cells had the predictable effect, where light could now be used to alter the previously random path of the constructs.
All this is interesting, but the article implies that it's some sort of important breakthrough or that it's nearly ready for some sort of purposeful use when all that's going on is some researchers are trying to construct synthetic organisms from stem cells like a kid building a lego kit, with results that are equally predictable.