Videos

Distributed Control of a Silicone Rubber Belt by Embedding Sensors, Actuation and Control

This video shows initial results toward the development of a novel class of soft robots "Chembots". On the long run, we are interested in combining chemical actuators, sensing and control into a novel kind of intelligent material that can exhibit drastic shape deformations and allow us to create more human-safe actuators than it is possible with the current paradigmn that relies on electrical motors, pistons and joints.

The video shows manual, open-loop and closed-loop control of a rubber belt that is equipped with 8 air chambers that can be inflated independently. For closed-loop control, we are using a microcontroller and a light sensor that help to localize each air chamber with respect to the surface and to determine the status of neighboring cushions.

Although the distributed control approach generates a sactisfactory forward gait, the experiment shows that solely controlling the sequence of actuation is not sufficient for making up for varieties in the material due to manufacturing.

The Distributed Robot Garden

Our long-term goal is to develop an autonomous green house consisting of autonomous robots where pots and plants are enhanced with computation, sensing, and communication. The network of robots, pots, and plants transforms energy, water and nutrients into produce and fruits. In this type of precision agriculture system water and nutrients will be delivered locally on-demand and fruit will be harvested optimally. Plants will drive the robots' activities in the garden using sensors to monitor their local environment conditions, a plant-specific model of growth for making predictions about the state of fruit, and interaction with robots for establishing an inventory of fruit. From an economical perspective, cultivation of specialty crops (such as fruits and vegetables) require a huge amount of manual labor and cultivation when compared with broad-land crops. This need has recently led to multiple initiatives in the United States (e.g. the Comprehensive Automation for Specialty Crops (CASC) program) and Europe (e.g. with in the scope of the 7th Framework program which aims at sustainable crop and forestry management, among others).

This project describes some first steps toward creating an autonomous distributed robotic garden as part of the undergraduate project course 6.084/086 taught at MIT during Fall 2008. The project was framed as addressing a grand challenge: to create a robotic gardening system. Solving the grand challenge required designing and programming robots to interact effectively and autonomously with the real world. We developed the class hardware infrastructure consisting of six robots with an iCreate base and a 4DOF arm with eye-in-hand configuration and an optional watering system and four cherry tomato plants, each with its own local sensing and computation packaged in an embedded computer. The robots an plants were networked together as a mesh network. The plants have the ability to monitor their soil humidity and issue watering requests. They also have the ability to database the location and color-level of the tomatoes. The robots have the ability to visit a specific plant to deliver water or to locate and grasp a tomato. Users have the ability to request tomatoes for salad. In response to user requests, the system decides which specific plants have the ripest tomatoes and assign parallel harvesting tasks to robots.

The Cow Whisperer - Towards Autonomous Management of Free-Ranging Cows

We demonstrate autonomous management of free-ranging livestock by manipulating individual as well as group behavior using intelligent sensor and actuator boxes without producing any noticeable signs of stress to the animals. Whereas herding is the classical method of animal control, wire fences are used today in most developed countries. Although wire fencing is effective for animal control, it is costly to build and maintain and does not foster flexible management, which is a key for spatially and temporally dynamic plants and animals. Autonomous animal management presents two challenges: first, algorithms must be developed for calculating desired trajectories for the animal's movement that satisfy both plant as well as animal ecology. Second, hardware is required that is robust, reliable and safe for the animals to wear for extended periods without human intervention.



This video depicts preliminary research into autonomous gathering of cows by melding electronics with animal behavior to accomplish control of cattle using a group of four mature crossbred beef cows on the Jornada Experimental Range (JER) operated by the US Department of Agriculture - Agricultural Research Service (ARS-JER). The first experiment was conducted in a corral on animals that were feeding; it focused on evaluating the effectiveness of several different non-voice aural cues to cause the animals to move to the left or right when the sounds were played in either the animal's right or left ear, respectively. Results show, that sounds alone are little effective: just playing sound did neither stopped the feeding nor did it initiate movement. A second experiment was performed to gather the animals back to the corral from a distance of approximately 1.5 km. Specifically, in a 122 ha (300 acre) arid-rangeland paddock four cows were autonomously gathered by playing the sounds the animals usually associate with manual gathering in both ears simultaneously. However, we believe that if the sounds had been administered directionally, the cows may have moved from the paddock to the corral in a more efficient manner. Future research using larger herds will focus on providing directional cues as well as determining how many animals within a group need to wear actuator boxes in order to control the entire herd. The hard- and software for these future experiments is currently under development at the Massachusetts Institute of Technology (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL).

Swistrack

Tracking of miniature robotic platforms involves major challenges in image recognition and data association. We present our 3-year effort into developing the platform-independent, easy-to-use, and robust tracking software SwisTrack, which is tailored to research in swarm robotics and behavioral biology. We demonstrate the software and algorithms abilities using two case studies, tracking of a swarm of cockroaches, and a swarm-robotic inspection task, while outlining hard problems in tracking and data-association of marker-less objects. Tracking accuracy of a moving robot with respect to camera noise and the calibration model are calculated experimentally. Its open, platform-independent architecture, and easy-to-use interfaces (MatlabTM, JavaTM, and C++), allowing for (distributed) post-processing of trajectory data online, make the software highly adaptive to particular research projects without changes to the source code. SwisTrack is publicly available on Sourceforge.net under the OSI Adaptive License and contributions from the robotics and biology community are encouraged.


N. Correll, G. Sempo, Y. Lopez de Meneses, J. Halloy, J.-L. Deneubourg, and A. Martinoli. SwisTrack: A Tracking Tool for Multi-Unit Robotic and Biological Systems. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 2185-2191, 2006.