PROCEEDINGS OF THE ROBOT GROUP ;^)

The Proto-Android Project
2nd Progress Report

June 23, 1993  -   Austin, Texas  -  David Santos

The Proto- Android (Pa) project has advanced by fits in recent months. This is a snapshot of progress. Detailed reports of topics only touched on
here are available.

David Hutchings has completed the machine work on the hips and shoulders. Lois and Adrian Herrera and I are doing the welded assembly of these
components as they come to us. We ready to freeze the pneumatic cylinder spec. in order to make the donation request approved by Bimba.

The waist is now thrashing around satisfactorily, driven by a custom battery powered compressor system. James Perez is finishing the second of two
pneumatic test-bed arms  for the control engineers to work with while Pa itself is built. J.P. Mason has been providing invaluable expertise to the pneumatic
design as has Vadim Konradi with timely research and suggestions.

Norman Annal has been creating splendid CAD drawings of Pa's mechanics and will be diagramming system schematics and architectural concepts developed
by Pa team members. 

Rand Grey, a highly regarded Motorola engineer referred by Alex, has offered to study Pa's control requirements. Considering the
difficulties, this a wonderful offer and we'll need to support his efforts as best we can.

Geoffrey L. Schecht, P.E. has provided valuable help to the control engineering effort. He is providing some onhand digital shaft encoders and researching
the best current options for various control components. 

Vadim is refining a mobile robot platform for systems development that will ultimately benefit many Robot Group projects including Pa. Wan-Yik Lee has
offered to share the A.I. based navigation and high level control system that he has developed, with Ben Kuipers at UT, to Vadim's platform and Pa. 

Wan is interested in overall architectures for intelligent robots and has invigorated group discussion in this area. Some of us will be attending the
AAAI- 93 Conference in Wash. D.C. to follow work in this area. Advanced mobile robotics has become an ever more prominent feature of the conference. Contact one of us to arrange participation. My mother lives in DC and can provide some accommodation.

Gilbert Andrade and I have been mocking up an artificial personae in FoxPro 2.0 (of all things). We are storing multi-media primitives in database tables and
writing rules for their expression. The demo incorporates limited speech recognition, speech out, and a graphical Face Space with over 3000 discrete
expressions that string together into interactive scripts. Things run very slowly now due to slow hardware and unoptimized code but should run well by the time
Pa is ready for an Artificial Personae System. The knowledge based structures needed to meaningfully integrate the personae are an ongoing preoccupation.

I'm continuing to build a data base of candidate resources for a state of the art android. It currently consists of a score of key data fields on a hundred
advanced systems for control, navigation, sensing, dexterity, intelligence, etc., some of which may be suitable for integration into Pa. The rationale for this exercise is that no small group like us can possibly duplicate the existing functionality in available systems. 

Finally; This project is absorbing group resources at a time when we are formalizing standards for project management and financial reporting.
Therefore I propose that Pa serve as model and proving ground for refining our institutional norms for documentation, cost effectiveness, technical excellence,
intellectual property, group/individual equity, review, etc..

The first progress report ended with an appeal for volunteers and that need is more acute than ever. Now is the time for the high level control freaks to
consider the potential and particulars of Pa as available development system.

Thanks for flying PolyCosmos. Join us again.