The following essay is reprinted with permission from The conversationan online publication about the latest research.
A honey bee’s life depends on successfully harvesting nectar from flowers to make honey. Deciding which flower is most likely to offer nectar is incredibly difficult.
To get it right, subtle clues about flower type, age, and history must be properly weighed — the best indicators that a flower can hold a tiny drop of nectar. Getting it wrong is at best a waste of time, and at worst it means exposing you to a deadly predator hiding in the flowers.
In new research published today in eLife our team reports on how bees make these complex decisions.
A field of artificial flowers
We challenged bees with a field of artificial flowers made from colored discs of cardboard, each offering a small drop of sugar syrup. “Flowers” of different colors differed in their likelihood of offering sugar, and also differed in how well bees judged whether or not the fake flower offered a reward.
We applied small, harmless paint marks to each bee’s back and filmed each visit a bee made to the floral array. We then used computer vision and machine learning to automatically extract the position and flight path of the bee. Based on this information, we were able to judge and accurately time every decision made by the bees.
We found that bees learned very quickly to identify the most rewarding flowers. They quickly judged whether to accept or decline a flower, but to their surprise, their correct choices were on average faster (0.6 seconds) than their wrong choices (1.2 seconds).
This is the opposite of what we expected.
Usually in animals – and even in artificial systems – a correct decision takes longer than an imprecise decision. This is called the trade-off between speed and accuracy.
This trade-off occurs because determining whether a decision is right or wrong usually depends on how much evidence we have to support that decision. More evidence means we can make a more accurate decision, but gathering evidence takes time. So accurate decisions are usually slow and inaccurate decisions are faster.
The trade-off between speed and accuracy is so common in engineering, psychology, and biology that you could almost call it a “law of psychophysics.” And yet bees seemed to break this law.
The only other animals known to beat the trade-off between speed and accuracy are humans and primates.
How then can a bee, with its small but remarkable brain, perform on par with primates?
Bees avoid risk
To break this question apart, we turned to a computational model, asking what properties a system would need to beat the trade-off between speed and accuracy.
We have built artificial neural networks capable of processing sensory input, learning and making decisions. We compared the performance of these artificial decision systems with that of real bees. From this we were able to determine what a system had to have in order to be able to make the trade-off.
The answer lay in giving different time-bound proof thresholds for “accept” and “reject”. Here’s what that means: bees only accepted a flower if they were at a glance Certainly It was worth the effort. If they had any doubts, they rejected it.
This was a risk-averse strategy and meant that bees might have missed out on some rewarding flowers, but it successfully focused their efforts on only those flowers with the best chance and evidence to supply them with sugar.
Our computer model of how bees made fast, accurate decisions correlated well with both their behavior and the known pathways of the bee brain.
Our model is plausible for how bees are such effective and fast decision makers. In addition, it gives us a template for how we can build systems, such as autonomous robots for exploration or mining, with these features.
This article was originally published on The conversation. Read the original article.