Honeybee colonies exhibit incredibly efficient and adaptive behaviors as a group, even though an individual bee is tiny compared to the world it lives in. Honeybee colonies regularly find and exploit resources within 2-6 km of their hive, adapt the number of bees exploring and exploiting multiple resources (pollen, nectar, water) based on the environment and needs of the colony, and can even recover when dramatic changes are made to their world. While much remains to be understood, biologists believe that many of these sophisticated group behaviors arise from fairly simple interactions between honeybees in the hive, as they share information and adapt their own choices. There seems to be no leader, no centralized authority, to coordinate the hive.

Achieving the sophistication of social insect colonies poses a number of challenges. It will involve the development of sophisticated coordination algorithms, that match the fairly simple and limited sensing and communication we expect in individual robobees. Just as with honeybees, the ability to leverage the colony as a whole will be critical -- for parallelism (exploration of large areas), energy efficiency (through information sharing and division of labor), and robustness (since individuals may fail or make errors). Especially since each individual robobee has strong limitations on the weight and power (and thus sensing/communication) it can carry.

At the same time, to manage swarms of robots (with thousands or more individuals) one cannot be managing single robobees. We will need programming languages and run-time tools that support a "global-to-local" approach. A key challenge will be the design and scalable implementation of macro languages, where goals can be expressed in terms of high-level objectives for the colony and where the underlying system translates objectives into individual bee decisions and re-optimizes as the world changes.

The RoboBee colony challenges are shared with many other fields in computer science -- for example multi-robot and robot swarm systems, distributed sensor networks, programming languages research, and even synthetic biology. Our colony team leverages expertise and knowledge in multiple disciplines, and we expect our methodologies to apply to many large-scale systems.

Some of our current efforts include

      (1) Karma Programming System and Stochastic control policies

      (2) Simbeeotic Simulation Environment

      (3) Heli-testbed Environment

      (4) Models of Honeybee Information-sharing

To read more about our current work, take a look at our publications  section.

You can also see videos of our work on our Robobees Colony Youtube Channel

For More Details on Specific Topics See Below:

A main area of research has been the design of programming languages

for controlling swarms of robotbess. We have developed two abstract

languages to start with. In the Karma language, one can specify a

flowchart of tasks that the colony must achieve with links that

represent conditions that trigger new tasks. The Karma system uses

information that comes back from individuals to adjust the allocation

of resources to tasks in a way that mimics the role of the hive in

real honeybee colonies. A different second, called OptRAD (Optimizing

Reaction-Advection-Diffusion), treats the colony of aerial robots as a

fluid that diffuses through the environment. Any individual RoboBee

uses a probabilistic algorithm to determine whether it will execute a

task based on the current state of the environment. OptRAD uses the

fluid model to efficiently reason at a macro-level about the expected

outcomes and adjust its behavior to adapt to new

circumstances. 

 

In addition to developing programming models we are also developing

testbed environments to study the design and operation of large-scale

swarms. One environment is the Simbeeotic simulator, that uses a

physics simulation to model scenarios and interfaces with a

vicon-based helicopter laboratory, to do mixed simulation-hardware

experiments. This expeditions project has also contributed to the

development of the Kilobot system: a collective of hundreds of robots,

each about the width of a quarter, that move by vibrating and that

communicate with other nearby Kilobots. We are using this collective

to test the efficacy of our programming languages and our mathematical

models of emergent behavior. Both systems have been made available

open-source, and the Kilobot robot has also been commercialized by the

educational robotics company K-team.