AI Robotics System Architectures PDF

4b System Architectures Can you make the “architecture” of AI for robots more tangible? What are the subsystems in a system architecture? I know a technical...

4b System Architectures Can you make the “architecture” of AI for robots more tangible? What are the subsystems in a system architecture? I know a technical architecture depends on the implementation, but what goes into a technical architecture? http://brrism.blogspot.com/2010/07/brrism11-edemocracyopen- datasocial.html © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 1 4b Specific Learning Objectives Be able to relate the functions in the canonical operational architecture to the 5 common subsystems Objectives Review Systems Classify a systems architecture as being either hierarchical, - 5 Subsystems reactive, or hybrid deliberative/reactive based on 1) the - Paradigms -Hierarchical relationship of the 3 AI robot primitives and 2) sensing handling -Reactive -Hybrid Technical Arch - evaluating Be able to draw the Hybrid Deliberative/Reactive System Architecture © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 2  Systems Architectures: 4b Outline Review Objectives Review 3 ways of generally organizing systems – 5 common subsystems Systems - 5 Subsystems - Paradigms -Hierarchical -Reactive – Why those 5? A historical evolution of Hierarchical, -Hybrid Technical Arch Reactive, Hybrid Deliberative/Reactive - evaluating – Contributions to canonical system architecture A little about technical architectures (technical) Summary © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 3  Recall: Types of architectures 4b [Levis, George Mason University] Objectives Review Systems operational architecture: describes what the - 5 Subsystems - Paradigms -Hierarchical systems does, not how it does it -Reactive -Hybrid Technical Arch - evaluating systems architecture: describes how a system works in terms of major subsystems technical architecture: implementation details, language © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 4  Types of architectures 4b Class will focus on these levels Objectives Review System Arch operational architecture: describes what the - 5 Subsystems - Paradigms -Hierarchical systems does, not how it does it -Reactive -Hybrid Technical Arch - evaluating systems architecture: describes how a system works in terms of major subsystems technical architecture: implementation details, language And the algorithms used at this level © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 5  The Most Abstract Canonical 4b Operational Architecture Objectives Review “Upper brain” or cortex System Arch Reasoning over symbols Interaction - 5 Subsystems - Paradigms (information) about goals Layer -Hierarchical -Reactive -Hybrid Deliberative Technical Arch “Middle brain” Layer - evaluating Converting sensor data into symbols (information) Reactive (or Behavioral) Spinal Cord and “lower brain” Layer Skills and responses © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 6  The Most Abstract Canonical 4b Operational Architecture Objectives Interaction Each layer has a Review System Arch - 5 Subsystems Layer - Paradigms -Hierarchical different style of -Reactive -Hybrid program organization Technical Arch - evaluating Deliberative Layer Not concerned with Reactive Layer details of interaction layer Behavioral Layer © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 7  More Tangible Canonical 4b Operational Architecture Objectives Review monitoring generating PRESENT+PAST+FUTURE, SLOW System Arch - 5 Subsystems World - Paradigms plan -Hierarchical -Reactive model -Hybrid selecting implementing PRESENT+PAST, FAST Technical Arch - evaluating PRESENT, VERY FAST sense act sense act sense act © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 8  monitoring generating DELIBERATION World plan model selecting implementing REACTION sense act sense act sense act Reasoning over information about goals: Promising results: Navigation, payload planning, contingency replanning Open issues: Multi-agent replanning, fault recovery & reconfiguration, reasoning over multiple failures Converting sensor data into information: Promising results: ATR, single failure health monitoring Open issues: creation of world models & situation awareness, monitoring & detection of new threats, exceptions, opportunities Behaviors, skills and responses 9  Types of architectures 4b [Levis, George Mason University] Objectives Review System Arch operational architecture: describes what the - 5 Subsystems - Paradigms -Hierarchical systems does, not how it does it -Reactive -Hybrid Technical Arch - evaluating systems architecture: describes how a system works in terms of major subsystems technical architecture: implementation details, language © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 10 4b System Architectures Provide… “right” subsystems – Focus on good software engineering Objectives Review – Modular (object-oriented) – Abstraction and cohesion System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive Libraries of algorithms and data structures within a -Hybrid Technical Arch - evaluating subsystem – Can pick from library to fit a specific niche Platform neutral © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 11 4b Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical WHAT ARE THE SUBSYSTEMS -Reactive -Hybrid Technical Arch - evaluating IN A SYSTEMS ARCHITECTURE? Via… A history of the three paradigms and example systems © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 12  Generally Accepted Subsystems 4b (or Objects) Navigation (Generating) – Planning about navigational goals, path planning Objectives Review System Arch Cartographer, World Model, World - 5 Subsystems Map (World Model) - Paradigms -Hierarchical – Everything to do with world models and -Reactive planning over maps -Hybrid Technical Arch - evaluating Planning (Generating mission, From Technology Development for Army Unmanned Ground Implementing, Selecting, Monitoring) Vehicles 2002, National Research Council Motor Schemas, Behaviors (executing motor commands) Perception, Sensing, Perceptual Schemas (executing sensor input) © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 13 4b Generally Accepted Subsystems Navigation (Generating) – Planning about navigational goals, path planning Objectives Review System Arch Cartographer, World Model, World - 5 Subsystems Map (World Model) - Paradigms -Hierarchical – Everything to do with world models and -Reactive planning over maps -Hybrid Technical Arch - evaluating Planning (Generating mission, From Technology Development for Army Unmanned Ground Implementing, Selecting, Monitoring) Vehicles 2002, National Research Council Motor Schemas, Behaviors (executing motor commands) Perception, Sensing, Perceptual Schemas (executing sensor input) © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 14  Describing Systems Architectures in 4b AI Robotics Subsystems can be thought of in terms of 2 Objectives Review attributes System Arch - 5 Subsystems 1. (relationship) How 3 building blocks, or robot - Paradigms -Hierarchical primitives, are arranged -Reactive -Hybrid Technical Arch 2. (content) How sensing is handled - evaluating This leads to 3 paradigms of system architectures for AI robotics – Hierarchical – Reactive – Hybrid Deliberative/Reactive © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 15  Recall: 4b AI Primitives within an Agent Objectives Review System Arch - 5 Subsystems SENSE PLAN ACT - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating LEARN © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 16 4b Hierarchical (1967) SENSE PLAN ACT Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 17 4b Hierarchical (1967) SENSE PLAN ACT Objectives Review System Arch - 5 Subsystems - Paradigms Control people hated -Hierarchical -Reactive -Hybrid Technical Arch - evaluating because didn’t “close the loop” AI people hated because monolithic Users hated because very slow © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 18  Hierarchical Paradigm for 4b Operational Architectures Objectives Advantages of Hierarchies (Albus, Mystel 01) – Natural way to organize Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive – Not intrinsically rigid -Hybrid Technical Arch - evaluating – Not intrinsically inefficient not the same as centralized planning priorities and goals are clear, therefore efficient © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 19  Hierarchical Paradigm: 4b Sequence of 3 Primitives Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical SENSE PLAN ACT -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 20  Hierarchical Paradigm: 4b Sensing is Centralized (global) SENSE PLAN ACT World model is a fused global data structure. It combines: 1. A priori representation 2. Sensed info 3. Cognitive understanding © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 21  Hierarchical Paradigm: Notable 4b Systems Architectures STRIPS/GPS (Nilsson) Objectives – Not used anymore, but did spawn planning Review System Arch - 5 Subsystems industry – Shakey - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 22 4b Shakey First AI robot Objectives Review System Arch Built by SRI - 5 Subsystems - Paradigms -Hierarchical (Stanford Research -Reactive -Hybrid Technical Arch Institute) for DARPA - evaluating 1967-9 Used Strips as main algorithm for controlling what to do © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 23  Hierarchical Paradigm: Notable 4b Systems Architectures STRIPS/GPS (Nilsson) – Not used anymore, but did spawn planning industry Objectives Review System Arch – Shakey - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch Nested Hierarchical Controller or NHC - evaluating (Mystel) – Mostly theoretical, oriented towards navigation – Divided into subsystems: mission planner, navigator, pilot, world model, low-level controllers © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 24  Nested Hierarchical Controller 4b (Meystel) Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 25 4b NHC Planner Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 26  Hierarchical Paradigm: Notable 4b Systems Architectures STRIPS/GPS (Nilsson) – Not used anymore, but did spawn planning industry Objectives – Shakey Review System Arch - 5 Subsystems - Paradigms Nested Hierarchical Controller or NHC (Mystel) -Hierarchical -Reactive – Mostly theoretical, oriented towards navigation -Hybrid Technical Arch – Divided into subsystems: mission planner, navigator, pilot, world - evaluating model, low-level controllers NIST Real-time Control System or RCS (Albus) – Integrated NHC and modified: sensory processing, world model, behavior generation, value judgment – Initially adopted by military in late 1980s, later versions try to retrofit to hybrid – Army UGV/XUV program © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 27 4b Demo III XUV Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating http://museum.nist.gov/exhibits/timeline/item.cfm?itemId=38 Experimental Unmanned Vehicle in action at Ft. Indiantown Gap. Photo courtesy of the Army Research Labs. Nov. 2001 © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 28  Demo III Control Hierarchy Section 10m PLANNER vehicle1 vehicle2 Vehicle 1m VEHICLE PLANNER communications plan AM plan RSTA plan Subsystem 5s COMMS PLANNER AM PLANNER RSTA PLANNER message list Driver Plan Gaze plan gaze plan Primitive 500ms DRIVER PLANNER GAZE PLANNER Velocity Plan Stereo Gaze Plan LADAR Gaze Plan Servo 50ms VELOCITY PLANNER F Wheels R Wheels F Steer R Steer F Wheel R Wheel F Steer F Steer 29  Disadvantages of Hierarchical 4b Organizational Architectures While hierarchies have advantages, relying on a Objectives world model creates problems Review System Arch – Bottleneck on processing, particularly for control - 5 Subsystems - Paradigms Alternative is to create layers or hierarchies within the world -Hierarchical -Reactive model to match other subsystems (Kaebling, Simmons) -Hybrid Technical Arch - evaluating c – World model requires extensive representation which leads to two major problems: Operates under the closed world assumption Frame problem In practice, implementations are planning-centric © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 30 4b Reactive (1986) SENSE ACT Objectives PLAN 1 Review System Arch SENSE ACT - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid SENSE ACT Technical Arch - evaluating 2 © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 31 4b Reactive (1986) SENSE ACT Objectives PLAN 1 Review System Arch SENSE ACT - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid SENSE ACT Technical Arch - evaluating Users loved it because it worked AI people loved it, but wanted to put PLAN back in 2 Control people hated it because couldn’t rigorously prove it worked © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 32  Reactive Paradigm: Two primitives 4b grouped into concurrent behaviors SENSE ACT SENSE-ACT PLAN couplings are “behaviors” SENSE ACT SENSE ACT Behaviors are independent, run in parallel, output is emergent © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 33  Reactive Paradigm: Sensing is 4b behavior specific (local) SENSE ACT SENSE-ACT PLAN couplings are “behaviors” SENSE ACT SENSE ACT Behaviors are independent, run in parallel, output is emergent © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 34 4b Reactive Robots RELEASER behavior Objectives Review System Arch - 5 Subsystems SENSE ACT - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating Behaviors (independent processes), released by perceptual or internal events (state) No world models or long term memory Highly modular, generic Overall behavior emerges © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 35  Reactive Paradigm: Notable 4b Systems Architectures Subsumption (Brooks) Objectives – No subsystems, just layers of competence Review AuRA (Arkin) System Arch - 5 Subsystems - Paradigms – Potential field implementation -Hierarchical -Reactive -Hybrid Technical Arch - evaluating Will cover in great detail later © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 36 4b Seriously, what can you without Objectives Review System Arch planning? - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 37 4b Robomow Behaviors? Random Objectives Review System Arch Avoid – Avoid(bump=obstacle) - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid – Avoid(wire=boundary) Stop Technical Arch - evaluating – Stop(tilt=ON) All active www.friendlymachines.com © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 38 4b My Real Baby Behaviors? Touch-> Awake Objectives Review System Arch Upside down & Awake-> Cry - 5 Subsystems - Paradigms -Hierarchical Awake & Hungry -> Cry -Reactive -Hybrid Technical Arch Awake & Lonely -> Cry - evaluating www.irobot.com Note can get crying from multiple behaviors © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 39  Reactive Robots Among Us 4b (not so obvious) www.disney.com Disney – Animatronics Objectives – Animation! Hunchback of Review System Arch Notre Dame - 5 Subsystems - Paradigms Web search engines -Hierarchical -Reactive -Hybrid Appliances Technical Arch - evaluating www.pfn.com/vision/default.cfm © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 40 4b Hybrid Deliberative/Reactive (1990) PLAN Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive SENSE ACT -Hybrid SENSE ACT Technical Arch SENSE ACT - evaluating © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 41 4b Hybrid Deliberative/Reactive (1990) PLAN Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive SENSE ACT -Hybrid SENSE ACT Technical Arch SENSE ACT - evaluating Control people hated it because no models at the lowest level, but are getting over it AI people loved it Users loved it © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 42  Hybrid Deliberative/Reactive: 4b Plan, then Sense-Act PLAN Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating SENSE ACT SENSE ACT SENSE ACT Plan, then sense-act until task is complete or need to change; Note movement towards event-driven planning rather than continuous © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 43 4b Hybrid Paradigm: Sensing Organization is Shared Deliberative functions Can “eavesdrop” Objectives Review Can have their own WORLD MAP/ System Arch - 5 Subsystems sensors KNOWLEDGE REP - Paradigms -Hierarchical SENSOR 3 Have output which -Reactive -Hybrid looks like a sensor Technical Arch - evaluating virtual sensor output to a behavior(virtual BEHAVIOR sensor) SENSOR 1 BEHAVIOR ACTUATORS BEHAVIOR SENSOR 2 © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 44  The Hybrid Deliberative/Reactive System Architecture: the Heart of the Canonical Operational and System Architecture Social Interface cartographer Performance World Mission & Problem World Planner Solving model World model model Sensing Sequencer Resource Mgr Effectors Behavior Behavior Behavior Behavior Behavior 45  Contributions of 3 Different Styles Managerial (division of Social Interface responsibilities looks like in business) – AuRA (Arkin) cartographer Performance World Mission & Problem World Planner Solving model World model model Sensing Sequencer Resource Mgr Effectors Behavior Behavior Behavior Behavior Behavior 46  Contributions of 3 Different Styles Managerial (division of Social Interface responsibilities looks like in business) – AuRA (Arkin) cartographer State Hierarchies Performance World Mission (strictly by time scope) & Problem World Planner Solving model World – 3T (Bonasso, Kortenkamp, model Firby, Gat) model Sensing Sequencer Resource Mgr Effectors Behavior Behavior Behavior Behavior Behavior 47  Contributions of 3 Different Styles Managerial (division of Social Interface responsibilities looks like in business) – AuRA (Arkin) cartographer State Hierarchies Performance World Mission (strictly by time scope) & Problem World Planner Solving model World – 3T (Bonasso, Kortenkamp, model Firby, Gat) model Model-Oriented (models Sensing Sequencer Resource Mgr serve as virtual sensors, Effectors multiple world models – Saphira (Konolige), TCA (Simmons) Behavior Behavior Behavior Behavior Behavior 48 4b Systems Architecture Systems architectures have evolved around functions and around paradigm Objectives Review System Arch - 5 Subsystems The hybrid deliberative/reactive paradigm “won” the - Paradigms -Hierarchical -Reactive paradigm wars -Hybrid Technical Arch – Most extensible and reusable - evaluating – More modular Hierarchical systems are still favored by DoD and NASA because of perceived efficiency and deterministic © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 49 4b Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating WHAT GOES INTO THE TECHNICAL ARCHITECTURE? © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 50 4b Technical Details… Algorithms such as Hough transforms, genetic algorithms, neural networks, path planning, fuzzy Objectives Review logic, scheduling, resource allocation System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive Data structures such as potential fields, perceptual -Hybrid Technical Arch spaces, polar plots - evaluating Algorithms for coordination and control of software modules such as island driving, vector summation, subsumption, voting, production rules Displays and interfaces such as natural language, gestures © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 51 4b Software-Service Flavor “Software as a service” extension to object-oriented programming Objectives – Modules/functions are used rather than owned Review System Arch – Independent of programming language - 5 Subsystems - Paradigms – Stateless -Hierarchical -Reactive -Hybrid Technical Arch - evaluating Practical ramifications – Libraries that subsystems draw from dynamically Contain logicially equivalent routines with methods to adapt them – Ability to store, distribute those libraries off-board (at a control center) Proxy processing, centralized control Use resources of other agents (group minds like Vinge’s Fire Upon the Deep) © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 52  Evaluating the Technical 4b Architecture support for modularity: – decomposition by functionality specified by the systems Objectives architecture Review System Arch - 5 Subsystems - Paradigms -Hierarchical niche targetability: -Reactive -Hybrid – Can be adapted to domains Technical Arch - evaluating ease of portability to other domains: – Infrastructure code should be reusable robustness: – Often neglected beyond reactive system tendency to “do the right thing” © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 53 4b Objectives Review System Arch - 5 Subsystems - Paradigms -Hierarchical -Reactive -Hybrid Technical Arch - evaluating SUMMARY AND ADDITIONAL THOUGHTS © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 54  Summary: 4b Architectures An architecture is the Big Picture of how to program an intelligent robot. Objectives Review System Arch - 5 Subsystems “Architecture” in AI robotics is often refers to - Paradigms – The operational architecture which captures the functionality -Hierarchical -Reactive Three layers -Hybrid – Reactive Technical Arch – Deliberative - evaluating – Interface – The system architecture developed by a researcher or research group Three paradigms for the system architecture – Hierarchical – Reactive – Hybrid deliberative/reactive – The technical architecture, usually a novel technique such as potential fields The operational and system architecture influences the technical architecture © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 55 4b Summary Can you make the architecture more tangible? – Have functions, the hybrid deliberative/reactive architecture which gives 3 layers What are the subsystems in a system architecture? – Navigation (Generating) – Cartographer, World Model, World Map (World Model) – Planning (Generating mission, Implementing, Selecting, Monitoring) – Motor Schemas, Behaviors (executing motor commands – Perception, Sensing, Perceptual Schemas (executing sensor input) – BUT THIS IS NOT ALL THAT IS NEEDED OR POSSIBLE What goes into a technical architecture? – Specific algorithms, control/coordination, knowledge structures © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 56 4b Next Lecture So far- AI is good with symbols (deliberative) and good with skills (reactive) and has ignored people (interface) - it’s the conversion of sensor data “in the middle” that is holding up progress Telesystems – How do we put people “in the loop” to compensate for the problems in converting sensor data to symbols, i.e., telesystem as a workaround? – How intelligent does a robot need to be for people who want to work remotely through it, such as for telecommuting, telemedicine, surveillance, etc., i.e., telesystem for remote presence? © 2019 Robin Murphy Introduction to AI Robotics 2nd Edition (MIT Press 2019) 57