Mayak, a magnitude -10 (minus ten) satellite; how is the large, delicate reflector expanded?

The July 14, 2017 NASA Spaceflight article Soyuz 2-1A launches with Kanopus-V-IK and over 70 satellites says:

Mayak is a three-unit CubeSat which was built by Tvoii Sektor Kosmosa – or “Your Sector of Space” – an independent, crowd-funded team of engineers in conjunction with the Moscow State University of Mechanical Engineering. Mayak – meaning Lighthouse – will deploy a highly reflective tetrahedral structure.

Each side of this structure has an area of four square meters, or 43 square feet. To ground observers, the satellite is expected to have an apparent magnitude of up to -10, making it one of the brightest objects in the night sky. The structure will double as a deorbit mechanism, hastening the decay of the satellite’s orbit.

Right now the orbit of the primary payload Kanopus-V-IK 2017-042A, 42825 should be a good guess for the rough predictor of Mayak passes, but once it deploys that giant reflector, things should change quickly.

How is this large reflector made from thin polymer film carefully expanded to and maintained in its full size and shape so that its faces are at least somewhat flat?

below: Screen shot from the YouTube video Mayak. No more space debris!

below: Mayak Reflector – Photo: CosmoMayak, From Spaceflight 101

below: Mayak Artists conception, From NASA Spaceflight

• It looks like the 4 edges of the reflector are held up with rollup measuring tapes. So deployment would be unrolling the measuring tapes. More details on their website cosmomayak.ru but it's mostly in Russian. Jul 16 '17 at 18:38
• @Hobbes that would certainly be fun to watch! The roll-up measuring tapes I've seen like to "roll up", not extend. Unless you pop open the case and then bowaaaannngggg they go crazy. The delicate film makes this a real challenge.
– uhoh
Jul 16 '17 at 18:52
• A measuring tape contains a spring that retracts the tape. Remove that, and the tape stays put, and can be rolled out with a motor. Jul 17 '17 at 7:28
• @Hobbes I see, so three carefully oriented tapes with the springs removed, one (or better yet three) motors could do it if the reflector film were folded in a clever way, and the film didn't have any tendency to stick to itself. This must be interesting to watch, I wonder if the whole mechanism first extends out of the 3U body before any of this happens. Somehow this reminds me a tiny bit of the challenges with the JWST films, although the solution is different.
– uhoh
Jul 17 '17 at 7:42
• Do we know what the observed angular width will be? Are we launching a new moon here?
– user7073
Jul 17 '17 at 17:08