Overview:
Recent research has developed sensor networks, where tiny devices with processors, memory, and wireless communication can be deployed to collaboratively sense and reason about the environment around them. Sensor nets are versatile enough to be thrown out into some outdoor environment, or can be integrated into offices and homes to create “smart buildings.” We have begun investigating the concept of Actuator Networks, where we enhance these passive sensing devices with distributed actuation, such as emitting different amplitudes and wavelengths of light or sound, to affect an environment. The general idea is to generate a “potential field” of attracting and repelling actuators.
We forsee a number of interesting theoretical and practical problems as we must not only develop new algorithms to determine the correct actuation for the desired goal in a general context, but we must also determine which modalities are most applicable for our subject to easily observe or be controlled by the potential field for our specific applications.
Emphasized Applications
Security
Examples we are currently investigating include using a pulse of sound to get a person’s attention for better facial recognition (which is currently very dependent on angle of the image) or using a bright light, unpleasant odor or annoying sound to coax people away from a region for issues of crowd control.
Natural Environments
We are experimenting with using modalitites such as light and sound to attract and scare animals in order to induce them to move to a desired location. For instance, we could actively guide birds toward a desired zone for closer viewing or herd animals animals.
Emergency Evacuation
We are exploring the use of action nets either pre-installed in "smart homes" or dynamically deployed by emergency personnel during search and rescue operations to help guide both those the search and rescue team and the trapped victims out of the building safely. In the firefighter scenario, the sensors (called motes) would sense heat and light up red or green in different amplitudes to shape a green corridor for rapid exit.
Related Material
Actuator Networks for Navigating an Unmonitored Mobile Robot. Jeremy Schiff, Anand Kulkarni, Danny Bazo, Vincent Duindam, Ron Alterovitz, Dezhen Song and Ken Goldberg IEEE Conference on Automation Science and Engineering (CASE). Washington DC. August 2008 [1.5MB .pdf].
Members
Profs: Ken Goldberg, Dezhen Song
PostDocs: Vincent Duindam, Ron Alterovitz
PhD Students: Jeremy Schiff, Anand Kulkarni
Undergraduates: Danny Bazo
UC Berkeley, Cornell University, and Texas A&M
Summer 2007 - Present
Example: Actuator Networks for Navigating an Unmonitored Mobile Robot
Abstract:
Building on recent work in sensor-actuator networks and distributed manipulation, we consider the use of pure actuator networks for localization-free robotic navigation. We show how an actuator network can be used to guide an unobserved robot to a desired location in space and introduce an algorithm to calculate optimal actuation patterns for such a network. Sets of actuators are sequentially activated to induce a series of static potential fields that robustly drive the robot from a start to an end location under movement uncertainty. Our algorithm constructs a roadmap with probability-weighted edges based on motion uncertainty models and identifies an actuation pattern that maximizes the probability of successfully guiding the robot to its goal. Simulations of the algorithm show that an actuator network can robustly guide robots with various uncertainty models through a two-dimensional space. We experiment with additive Gaussian Cartesian motion uncertainty models and additive Gaussian polar models. Motion randomly chosen destinations within the convex hull of a 10-actuator network succeeds with up to 93.4% probability. For n actuators, and m samples per transition edge in our roadmap, our runtime is O(mn6).
Related Figures
 |
| An actuator network with sequentially activated actuators triplets (shown as squares) driving a mobile robot toward an end location. The robot is guided by creating locally convex potential fields with minima at waypoints (marked by xs). |
 |
| Example of a simulation of the actuator-networks algorithm with 8 actuators (denoted by circles) and incenters formed by sets of three actuators (denoted by *s). The actuators are placed randomly on the border of a square workspace, and work in sets of three to exert forces on the robot to move it from one incenter to another. Therefore, the incenters of all possible triangles between them form vertices in a roadmap with edges containing the probability of successful transition by activation of an actuator triplet. The left hand figure illustrates the placement of actuators, triangles of three actuators, and incenters. The right hand figure shows the roadmap and an example path (thick line) through the network. |
Funding:
|
NSF Science and Technology Center, Team for Research in Ubiquitous Secure Technologies, NSF CCF-0424422, with additional support from Cisco, HP, IBM, Intel, Microsoft, Symmantec, Telecom Italia and United Technologies. |
| CONE Project NSF Award 0534848/0535218 Robotics and Robust Intelligence Program Division of Information and Intelligent Systems Directorate for Computer Science and Engineering National Science Foundation |
|
| AFOSR Human Centric Design Environments for Command and Control Systems: The C2 Wind Tunnel, under the Partnership for Research Excellence and Transitions (PRET) in Human Systems Interaction. |
