Bees bounce into things rather than navigate around them. How robobees could do the same

How Harvard's Robobee nanodrone project is using the learning from real bees to improve the technology. Flexible wings that can handle collisions don't require as advanced sense-and-avoid skills. 

Views: 719

Comment by Gary McCray on May 14, 2016 at 3:18pm

In this world scale is everything, now all I have to do is figure out how to stuff in a 3D vision system, a 4K camera, a few thousand Pascal GPUs and a tiny battery or maybe better still SNAP generator.

Can't wait till the FAA starts trying to regulate bee sized drones.

Comment by Maxime C on May 14, 2016 at 3:18pm

According to Nature's article about small drones, bees also use their compound-eyes called ocellis to get sereophotogrammteric optic flow.

That is such an interesting topic, altough i personally think lidars will know more success, since contrary to insects our drones' brains will have the capability to process such accurate and important data.

Comment by HeliStorm on May 14, 2016 at 5:57pm
Nature has engineered some very unique unmanned aerial systems. Self-guiding, self-fueling, self-manufacturing nano aircraft. It shows what is possible when you have billions of years to trial and error.
Comment by RichWest on May 14, 2016 at 6:34pm

Interesting future, no doubt.  Really interesting are those that see and then attempt to solve the puzzles of nature.

Comment by Digital Wings on May 15, 2016 at 5:44am
very interesting!

Developer
Comment by John Arne Birkeland on May 15, 2016 at 8:48am

@Gary. Ironically once we have safe to use bee sized drones with the capabilities you mention, it is also the moment the current privacy fears become a real problem.

Comment by Gary McCray on May 15, 2016 at 9:39am

@ John, I really think that ship has already sailed.

Comment by HeliStorm on May 15, 2016 at 4:46pm
I have been giving this some more thought, and I think there is a real potential for systems able to bounce around obstacles. One of my interests is search and rescue applications, but in particular the usage of smaller systems in wooded areas, under canopy, to allow for quick, "over the hill," reconnaissance. If a small team can quickly visualize an out-of-reach area with an aerial system, they will be able to manage their resources, potentially speed up efforts by ruling out areas that might take longer to reach on foot, and can protect responders better by keeping them out of dangerous situations where it is not needed.

The issue with flying in wooded areas is the trees themselves, which prevent good GPS, and are obstacles that could easily knock a aerial system out of the air. A pilot flying in these areas would potentially need to be an expert FPV pilot, capable of quickly dodging the wayward branch. A better option would be some level of autonomy, where an aerial system could avoid these dangers. Progress is being made on obstacle avoidance, but it is still in larger systems, and expensive.

Now, imagine instead a system that allows the pilot to set a general course of travel, and the aerial vehicle goes that way until told otherwise. It seems it would be easier to program a system to self right its course based on IMU and compass information after a collision. Put this into a frame designed to take a bump, like some of the smaller proximity FPV setups that have fully enclosed blades, or any of the globe-like frames I have seen too which essentially roll around the internals, and you have your system.

Alright, lets do this!

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

Groups

Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service