TUTORIAL SESSION

Biologically-Inspired OOD / OOP for Robotics and AI

Dr. Paul Frenger
A Working Hypothesis, Inc., USA
pfrenger1@comcast.net

Abstract

The author, a physician, roboticist and artificial intelligence (AI) researcher, demonstrates the techniques of object oriented design (OOD) and programming (OOP), using a biologically-inspired model: a meter-tall android robot. Dr. Frenger created a multi-core, multiprocessor, multitasking robot control system based on IEEE 1275-1994 for his robot ANNIE (acronym for "android with neural network, intellect and emotions"). ANNIE has an artificial eye with machine vision, speech recognition / speech synthesis, an artificial neural network and a dexterous hand. ANNIE's nervous system contains a knowledge base, schema-script executive, long and short term memories, synthetic emotions, personality preferences and limited self awareness. The nervous system model was extended to emulate normal and abnormal pediatric growth & development, pain and addiction (opium narcotics), hormone action (epinephrine, oxytocin), the mammalian fear mechanism, the endocannabinoid system, a genetic-based disorder (autism), and immune system diseases (fibromyalgia). A simian version of ANNIE for space adds a grasping hand-foot useful in weightlessness. How to decompose complex biological organ systems and neural functions into subunits which can be represented in hardware and software is shown in detail. This system is mostly data-driven and relies on program language-independent byte-code. "Structured English" pseudocode is used for many software examples. System software discussed ranges from low-level hardware drivers to high-level behavioral scripts.

Objectives

Attendees will learn techniques of object oriented design (OOD) and programming (OOP), using a biologically-inspired model: a meter-tall android robot. Functional decomposition of complex biological organ systems and neural functions which can be represented in hardware and software is shown in detail. This system is mostly data-driven and relies on program language-independent byte-code. Attendees will learn how to use "Structured English" pseudocode to produce software. Integrating hardware objects with software objects in an extensible plug-and-play android robot will give attendees valuable skills which are applicable in a variety of real-world applications, not just robotics.
This tutorial covers material for which significant validation exists. This includes accepted standards (IEEE 1275-1994; IEEE 1451.4; IEEE P996.1 draft, June 1996: ANSI X3.215-199 Forth language, and others). Some components are being used in non-trivial applications and/or real-world settings, and were presented in numerous publications in ACM, ISA and IEEE journals, and others (both peer-reviewed and non-peer-reviewed).

Timeline

The tutorial will mainly be a slide-based lecture, but with audience participation including some group problem-solving sessions and examination / operation of some robot components. Participants may use their laptops for Internet horizontal enrichment but a computer (other than the author's) is not required.
Major topics and time allocations include: a. introduction to object oriented design and object oriented programming (10 minutes); biologically-inspired systems (10 minutes); functional decomposition of complex problems (10 minutes); introduction to structured English pseudocode (10 minutes); top-down design and bottom-up programming (10 minutes); network solutions (10 minutes); using language-independent bytecodes (10 minutes); implementing robot operating systems and plug-and-play mechanisms (15 minutes); anthropomorphic robotics (sensory, motor) (20 minutes); introduction to artificial intelligence (20 minutes); emulation of the human cortex via a computer (20 minutes); enhancements to body system emulation (20 minutes); summary and wrapup (15 minutes).

Background Knowledge Expected of the Participants

Attendees should have a basic interest in biologically-inspired systems, robotics, artificial intelligence and object oriented programming techniques. They should be able to write a program in at least one computer language. Language-free "Structured English" pseudocode examples will be used in many places. This tutorial is intended to contain elements useful to and of interest to: Beginners, Intermediate, and Advanced level participants. Although aimed principally at students / educators, this tutorial will be of interest to a variety of researchers and practitioners in robotics and AI.

Qualifications of the Instructor(s)

Paul Frenger is a member of ACM, a Senior Member of IEEE and a practicing medical doctor in Texas working with computers professionally since 1976. He published over one hundred forty articles in the bioengineering and computer literature, served as editor for the ACM SIGForth Newsletter (five years), as associate editor for ACM Sigplan Notices (thirteen years), and has three computer patents. He is currently active in artificial intelligence, robotics, prosthetics, bioengineering and space science. Paul assisted US Air Force Research Labs (AFRL) scientists developing testing systems for non-lethal weapons (2004-2005); was elected first Chair of the Houston Chapter of the IEEE Computational Intelligence Society (2007); and twice served as General Chair of the Workshop on Automation and Robotics, NASA (2008-2009).
A former US Air Force Lt. Colonel, Dr. Frenger has been a member of the Rocky Mountain Bioengineering Symposium (RMBS) since 1983; the Canadian Medical and Biological Engineering Society (CMBES) since 1997; the Association for Computing Machinery (ACM) since 1990; the Institute for Electrical and Electronic Engineers (IEEE) and the Engineering in Medicine and Biology Society (EMBS) since 1989; the Instrumentation, Systems and Automation Society (ISA) since 2006; and MENSA since 1978. His biography is listed in: Who's Who in Medicine and Healthcare, Who's Who in Information Technology, Who's Who in Science and Engineering, Who's Who in the World, and others. He received the Air Force Commendation Medal (1978), two service awards from ACM (1992, 1999), one from IEEE (2008) and the RMBS President's Award (2004). His computer and automation consulting company, A Working Hypothesis, Inc, was chartered in 1982. He hosted 15 hourly medical talk-radio show episodes on KPRC Radio, Houston TX from 1992-1993.
The author's ACM publications are accessible at: http://www.acm.org/portal, search for "Frenger P". The author's National Library of Medicine publications are accessible at: http://pubmed.gov, again search for "Frenger P".
At ACM OOPSLA '95 (Austin, TX) the author conducted two 45-minute demonstrations of "A Distributed Control System for a Robotic Hand/Arm Model" which were both SRO attended. At OOPSLA '99 (Denver, CO) the author presented a popular poster on the "Objects in ANDROID.FORTH", which resulted in a book offer to the author from McGraw-Hill. Over the years since then, components of this proposed presentation were delivered at a variety of conferences (NASA, Mayo Clinic, Oxford University UK, US Air Force Academy, Schloss Dagstuhl Germany, Canada and other locations in the US) as well as within the pages of ACM Sigplan Notices and ACM SIGForth Newsletter. The author is a popular presenter and has often been an invited speaker.