Hmmm
Not a bad idea, within reason, so long as we don't end up having to worry about kessler syndrome around every major body in the solar system.
NASA has revealed it is working on CubeSat Application for Planetary Entry Missions (CAPE), a plan to use small satellites to explore the solar system. As Lori Keesey of the agency's Goddard Space Flight Center details, cubesats are proving very useful around Earth. So useful that the 10cmx10cmx10cm craft could be used to …
".... what battery technology they use for the very cold environment of space?"
A number of non-battery technologies are applied to change the spacecraft's environment to be more battery-friendly: insulation, radiators, and active thermal control. This both keeps a probe warm when it's cool, and cool when they overheat due to sunlight and internal operations. One of the most potent features used to keep spacecraft warm (or too warm) is space itself: spacecraft operate in the universe's largest Dewar flask.
The end results are interesting: For example, as I recall, the Voyager probes out in BFE run at about 70-80F to this day. And remember the ESA was worried that Philae was going to overheat on a comet that was, what, 3.5AU from the sun at intercept?
The batteries used in that environment vary by year and mission. Some are hardened, long-life versions of everyday battery technology, while others have only found use in space. Common spacecraft examples include nickel-cadmium (favored for 40 years in space missions); nickel-hydrogen for rechargeable applications (the ISS and Hubble both use nickel-hydrogen); some nickel-metal hydride batteries; and lithium-ion (used on Spirit and Opportunity, and heading to the ISS in 2016).
Spacecraft sometimes dip into unusual batteries for short-term applications. Single-use probes (like Huygens and the Galileo Atmospheric Probe) like to use thermal (molten salt) batteries, which are commonly found in single-use military applications like missiles.
https://youtu.be/giks4YQDMU0
Gerry Anderson got prophetic again 30-something years ago. If a martian space-witch turns up as part of the project, be very afraid (or send for Sergeant Major Zero ).
"It's a 10cm cube & they need to test its aerodynamics?"
It's a 10cm cube with differing centers of mass depending on the exact contents of the Cubesat that might thus twirl or fall in a manner different from a homogenous cube. It is also a 10cm cube that might be placed in an aeroshell.
A few years ago JPL looked at flying at flying as a secondary payload on comm sat launches to GEO. This (in theory) gave launch opportunities to somewhere (Moon, Mars, Venus, Mercury, anywhere really) on almost every launch.
The big benefit of this approach (in principal) is that it turns a once a decade mission to somewhere, to maybe once every 12-18 months, ideally using the data collected to drive the next generation of probes.
The challenge with these systems is to choose experiments that make the best use of this ability to sample multiple sites. It's also more likely a good idea to focus on 1 cubesat/1task
That is a sea change in how missions are planned or partitioned. 1 mission per cube sat or 1 mission per swarm of cube sats?