The acknowledged affirmation of remote-controlled “drones-bees” by Netherlands’ Delft University of Technology (TU Delft) has been making news since Tuesday. But the buzz surrounding the drone’s applications could be much hyped, according to at least two absolute estimates by biologists.

Many accepted news websites appear that the university’s drone-bees could auspiciously alter real bees in pollination if they go extinct, but this might be optimistic. First, it’s not economically achievable to mass aftermath the robot insects. Second, there’s a flaw in compassionate the science of pollination.

The Netherlands is one of the better exporters of food and agronomical articles in the world. While bees abide to blend 80 per cent of edible crops grown there, their citizenry is under connected threat from pesticides. Some experts warn that all 360 breed of bees in the country are endangered.

Enter the drones developed by TU Delft researchers. The robot bees have a wingspan of 33 cm, weigh 29 grams, can flap their wings 17 times per second, and accomplish enough lift to stay airborne. They ascendancy their flight by making minor adjustments in the wing motion. The university’s press absolution states that these appearance allow the drone bees to hover on the spot, and to fly in any direction. But they are not quite the same as actual bees.

The Guardian appear that the automatic insect can only fly for six minutes, or 1 kilometer (0.6 mile) on its accepted battery. They’re also about 15 times larger than actual bees, although the university plans to cut the size down in the future.

Speaking to the Telegraph, Mat?j Karásek, a researcher alive on the activity said “The goal is to have a device that could blend flowers as it may be that bees will die out…if you can brainstorm a swarm of these robots flying around a barn or greenhouse, they would very safe to work around as they are so small and light.” He added that they hope to eventually mass aftermath the drone so that it can be used worldwide.

While the drone bees are indeed futuristic, using them for large scale pollination may not be a absoluteness anytime soon. First, accomplishment these robots on a scale large enough to alter real life bees may not be an able solution.

Discussing the achievability of using drone-bees in large scale pollination, David Goulson, assistant of analysis at the University of Sussex and architect of the Bumblebee Conservation Trust in the United Kingdom, wrote a blog post last year. His back-of-the-envelope adding appear that if it cost just a single penny to build a single drone-bee (which is a ridiculously absurd assumption), it would cost £32 billion ($45 billion) to alter every honey bee in the world. He wrote:

While I can see the bookish absorption in trying to create automatic bees, I would argue that it is awfully absurd that we could ever aftermath article as cheap or as able as bees themselves… Consider just the numbers; there are almost 80 actor honeybee hives in the world, each absolute conceivably 40,000 bees through the spring and summer. That adds up to 3.2 abundance bees. They feed themselves for free, breed for free, and even give us honey as a bonus. What would the cost be of replacing them with robots?

Even if we cut down on ambition, and go by the words of Karásek that the drone-bees can help in pollination within greenhouses, there’s addition major hurdle – botany.

Manu Saunders, a association ecologist based at the University of New England, Australia, argued that acknowledging actual pollinators would be a more viable long term solution. One of the main problems she points out, is that pollination is a accustomed abnormality and its success often depends on abundant factors that are too circuitous for a drone to emulate. As she states in her blog post appear in April::

Successful pollination often depends on where the pollinator visited beforehand: pollen from other plant breed can reduce pollination success, pollen from other varieties of the same plant breed can affect pollination in altered ways, and bacilli or fungi that artlessly live in the nectar of other flowers can affect reproduction of addition flower. Managing these accustomed interactions is beyond the scope of a drone.

Saunders added that many crops crave assorted visits by altered types of pollinators (insects, animals, birds etc.). Only with assorted attempts, she noted, there is a chance of the right pollen auspiciously agriculture the flower to aftermath fruits or seeds. Sometimes there can even be adorning deformities in fruits and vegetables if pollination isn’t done right.

There are hoards of questions still abstruse biologists about pollination, she added: the time an insect has to spend on each flower; their most suited concrete behavior on the flower; the optimal number of pollen grains they must carry; the best time of the day for pollination; the ideal temperature and clamminess for pollination, and so on.

Additionally, every plant breed has unique flower types, and each of these are tuned for acknowledged pollination with altered insect species. In many cases, the flowers and their pollinators have gone through millions of years of accommodating evolution, and in the action have become mutually optimal for pollination. Drones may not just be able to break into this well choreographed dance of co-evolution.

Saunders assured that drone projects aiming to mimic bees could be fun to work on, but not as applied solutions to large scale pollination.

These pollination drones will be fun toys, and they could also be useful educational tools. They might even be destined for a life on Mars. But, no fear, they wouldn’t last long on most farms on Earth.

It must also be noted that neither the analysis paper appear on the drone project, the project’s official website, nor the press absolution from Delft University of Technology mentions the abeyant use of the drones as pollinators.

Read next: South Korean regulator warns of austere fallout if ICO ban lifts