1. Sounds too good to be true 2. Can it transfer from lab to commercialisation? 3. Would it overcome resistance from established systems, if so?

One of the most interesting articles about new tech I have read in a long, long time. Sadly, they brought up metal fatigue but then never really answered the question. A simple "it can run x% as many cycles as a traditional air conditioner" would have made the technology sound usable.

Sounds absolutely revolutionary if they can make the material last! Very cool! (pun intended). This would be great for van dwellers, a silent air con system.

Not quite perpetual motion, but with that huge ratio of cooling power to mechanical energy needed to operate the thing, can they run a heat engine off the difference between the heated and cooled air streams and use that to turn the wheel?

And it only cost 1,000 times more than a conventional system, and probably will use more electricity than a conventional system LOL.

Useless in high humidity environments.

guzmanchinky; I think you missed something. This system will require TWO fans, or one larger fan ducted in two directions, so its unlikely to be silent, and as it will be blowing across wires, that may make even more noise. However, any system that requires no polluting chemicals to be disposed of when the system becomes redundant must be good. There are thousands of old scrap fridges waiting to be decommissioned and their toxic refrigerants neutralised, but as they deteriorate, the refrigerants tend to escape anyway, and they are a serious threat to the ozone layer. Lets hope it reaches production, in an affordable form.

Not only is there no mention of useful cycles for the material, but does the effect weaken with age, (similar to recharge cycles on batteries),? But the concept is very interesting and most welcome as air-conditioning alone is one the world biggest users of energy. We certainly need alternatives to the current tech. Something worth watching I think.

I don't believe it! Heat pumps have a Coefficient of Performance of greater than 1.0 (100%) because they are not making heat; they simply move existing heat against it's natural gradient. What is proposed here does not involve moving heat but making it by bending the wires. The basic laws of energy conservation are being broken. Where is that heat coming from other than than the work put into bending the wires? Where does all the extra "free" energy come from?

I'm more than a bit skeptical here, but I will explain why. I've actually worked with NiTiNOL, years ago at LBL (Lawrence Berkeley Lab) on several different variations of motors, and various research platforms characterizing the thermal and crystallographic state phase transition properties of this alloy. As for the fatigue aspects of the alloy, MTBF (meantime between failure) can in fact extend through many millions of state phase transition cycles. But that's not the limitation here . . , The limitation is the efficiency of mechanical work vs. thermal gradient produced by the material as it is mechanically forced into its respective state phase transition cycles. The amount of thermal gradient produced by mechanically cycling the material is no where even close to the amount of mechanical energy required to induce that state phase change. We tried this, many times, with myriad different configurations, annealing thermal set points and variations of the nickel / titanium ratios in the alloy. In other words, the material can be mechanically cycled with external force to produce heat or absorb heat (cooling) as the contortion of the material induces the aforementioned state phase change, or . . . the material can be subject to alternate exposure to hot and cold environments, which causes a state phase change in the alloy (thus performing mechanical work).
The success we did have consisted of solar powered mechanical motors configured with NiTiNOL wire loops that cycled between immersion in hot and cold water vessels, forcing the contraction and expansion of the wire loops around geared bobbins. The concept was to create solar powered mechanical motors that operated on the thermal gradient between hot (solar heated) and cold (from wells) water.
NiTiNOL is commercially available in a variety of alloy ratios and annealing thermal set points to further customize the state phase transition vs. temperature thresholds, but the actual process of trying to cool air by cycling it past NiTiNOL as it is transitioning into its cold phase would require a phenomenal amount of energy spent, not to mention an extremely large surface area vs. CFM flow rate of air to actually produce effective cooling. I could be convinced otherwise, but I would very much like to see some hard data, engineering notes, or even 3rd party peer reviewed research on this before getting too excited about this general idea.