Contact the maintainer of this page.(Remove the NO_SPAM before sending)
General InfoThe Zaza project is part of an ongoing effort to make real robots more accessible to people. Much has been done in the popular press and the entertainment industry to obscure the true facts about robots, and in many respects this had led to a general mistrust of technology, and it's anthropomorphized embodiment: the mobile robot. Adding to this mistrust is a lack of information about robotics accessible to the general public. Most of the truly fascinating work is done in research institutions, and when presented publicly is often done so outside the level of comprehension of those not directly involved with robotics research. Some noted exceptions are the NASA Sojourner, which has generated a fantastic level of public interest and spurred many young people to get involved with robotics. Other projects such as Sebastian Thrun 's work on Minerva and the Nursebot are prototype projects with the aim of having mobile robots serve a direct and practical purpose for society today. Programs featuring Zaza allow museum visitors to interact with the robot and have trained staff and volunteers on hand that can provide answers to any questions during demonstrations.
Media CoverageOn January 4th 2001 Zaza was featured in a segment on robotics done by Good Morning America. Video of the segment is currently being hosted by Geocities, and can be found on this page .
PeopleMichael Drennan (Technology Developer)- Tech Museum liaison, Windows code development
Rich Turner (Engineering Director)- Technical support assistance, and intra-museum coordination
Steve Bischoff (Volunteer)- Program facilitator, battery management advisor
Robbie Stone (Volunteer)- Program facilitator, hardware interface programming
Brian Rudy (Volunteer)- Robot programming, Beesoft application development, hardware development
Andy Rubin (Sponsor)- Hardware donor
Sebastian Thrun - Beesoft developer, Advisor
Project DetailsThe project is currently in phase two of four phases. The phase descriptions and their status are shown below.
- Simple obstacle avoidance using enhanced version of wander demo with laserServer support, button-driven staff interface. Staff facilitated program. (complete)Status- Software written and tested. Acrylic dome cover for upper turret has been mounted. Button-based demo startup/shutdown written and tested. Simple web-based monitoring of program execution, startup/shutdown. Tech Programs management will provide training timeline and initial deployment schedule. Deployed Summer 2001
- wander-laser based simple obstacle avoidance program. Simple behavior/voice response via Java face applet. Program deployed September 29, 2001
Status- the colliServer wander program has been written and tested, Mike's program renders a mouth which is displayed on a small LCD monitor mounted on the camera bar of the pantilt unit. Speech is generated via the same program. Speech cues are generated randomly, and via a web-based interface. Staff facilitated for safety. Deployed September 2001.
- colliServer/LOCALIZE-based wander program. Detection and interaction with visitors. High-level obstacleServer collision avoidance in mapped areas requires only one facilitator.
Status- Cues for what is said are sent from an enhanced version of reaction using the DETECTION package to track people. It should be able to detect crowding, someone dodging in front of the robot, etc... Maps of all areas where the robot will be operated are necessary. DETECTION uses LOCALIZE to find discrepancies in map for people detection. Video streaming and limited control via web. Staff facilitated for safety. Deployment TBD.
- Web-based and intra-museum interactive tourguide allowing web visitors to visit exhibits and hear and explanation of what they are seeing (all galleries mapped, LOCALIZE used for position estimation)Status- New maps must be generated for the upper, lower, and ground floors. Several major exhibit features have changed in the upper level that will require re-mapping of the Innovations gallery. Final testing prior to deployment in-progress.