4 a2
Introduction
AI Robotics Overview
Exploring the necessity and hierarchy of intelligence in robots.
Learning Objectives
Definitions and Design Process
Apply Levis' definition of architectures in intelligent robot software organizing.
Familiarize with layers in a canonical operational architecture: behavioral, deliberative, and interface.
Describe the four primitives of robotic intelligence: SENSE, PLAN, ACT, LEARN.
Recognize vulnerabilities leading to failures in canonical operational architecture.
Architectures Overview
Types of Architectures
Operational Architecture: Describes what the system does at a high level.
Systems Architecture: Details how major subsystems of a system work.
Technical Architecture: Covers implementation details including programming languages.
Importance of Organizational Structure
Software Architecture
An overall design style termed "architecture" that dictates control system organization and interaction among components.
Supports software organization and influences problem-solving approaches.
Processing Architecture
Three Tier Hybrid Architecture
Characteristics of operational, biological, and process architectures.
Primitives
Enable robots to SENSE, PLAN, ACT, and LEARN.
Reaction Type Framework
Behavioral Robotics
Instinctive SENSE-ACT couplings without a planning phase.
Many behaviors acting on concurrent stimuli.
Fast and modular but limited predictive capability and coordination challenges.
Learning and Adaptation
LEARN
Incorporates diverse forms of learning, capable of enhancing the architecture.
Functional Considerations
Functional Observations
Factory automation functions like generating, implementing, and monitoring require integrating primitives.
Struggles exist in knowledge structures and implementation due to monolithic routines.
Layered Programming Benefits
Programming Styles
Different layers allow for specialization in programming languages suited for each task type, improving modularity.
Enhances reusability and adaptation capabilities, accommodating evolving requirements.
Decision Making and Models
Planning Horizons and Time Scales
Establishing the type and detail of models (local or global), dictates algorithm updates and decisions.
Conclusion and Future Considerations
Autonomy Limits
Questions arise on the trustworthiness of robot autonomy and its interaction critically with human oversight.
The evolving programming capabilities aim to enhance robot autonomy and improve function efficiency.