Design and Virtual Prototyping of Human-worn Manipulation Devices Peng Song GRASP Laboratory University of Pennsylvania ASME DETC99/CIE-9029 GRASP Laboratory.

Design and Virtual Prototyping of Human-worn Manipulation Devices

Peng SongGRASP Laboratory

University of Pennsylvania

ASME DETC99/CIE-9029

GRASPLaboratory

Venkat KroviMechanical Engineering Dept.

McGill University

Richard MahoneyRehabilitation Technologies Division

Applied Resources Corporation

Vijay KumarGRASP Laboratory

University of Pennsylvania

MOTIVATION

Mass production fixed automation flexible automation

Mass customization Agile manufacturing

speeds up the process of going from concept to production

Customized design and manufacture Human-worn products (helmets, hearing

aids, eye-glasses, wearable computers, ...) One-of-a-kind products: Product volume is 1

(3-day cars, assistive devices, ...) Human-worn manipulation assistive devices

In 1992, the cost of caring for quadriplegics was 11 billion dollars.

10,000 new spinal cord injuries every year (55% are quadriplegics, 58% between 16-30 years old)

Population aging is increasing the number of people requiring physical assistance (est. 60 billion dollars annually.

Annual cost to assist in Activities of Daily Living is around $80,000.

WHY ASSISTIVE DEVICES?

WHY CUSTOMIZATION?

Variability exists in user needs across the population. Products must be designed and customized to match these individual user needs.

The device-user customization ensures comfort of user and enhances performance of the device.

HUMAN-WORN MANIPULATION DEVICES

HUMAN-WORN MANIPULATION DEVICES

GOALS

Identify and investigate the component technologies required for designing, customizing, virtually prototyping and finally fabricating human-worn manipulation assistive devices for the motor disabled.

Present a unified design environment which integrates these component technologies and aids the designer in shortening the design cycle.

The design-customization-integration process can be extended to many classes of human worn products.

KEYS TO CUSTOMIZATION

1. DATA ACQUISITIONMeasurement of the human user, the task, and the environment.

2. DEVICE DESIGN AND OPTIMIZATION

Mechanism synthesis (generating the desired “output” motion/force from the specified human “input” motion/force), CAD modeling.

3. VIRTUAL PROTOTYPING AND EVALUATION

Geometric and dynamic modeling of the human user, the designed product, and simulation of the human using the product prior to rapid fabrication.

1. DATA ACQUISITION

GEOMETRY CAPTURE

in 3-D studio

CUSTOMIZED MODEL

MOTION CAPTURE

in 3-D Studio

1. DATA ACQUISITION: Measurement to models

Solid models generated from image data using Multi-camera, multi-pose measurements Cyberware 3D scanner

Meshed Solid models for CAD (Pro/Engineer) CAM (CNC machining) FEM/FEA

Kinematic and dynamic models for Virtual prototyping Analysis and simulation

Provides important tools for Re-design Customization One-of-a-kind prototyping Reverse engineering

DESIGNER

AUTOMATION

2. DEVICE DESIGN AND OPTIMIZATION

COUPLINGCOUPLINGINPUT DEVICEINPUT DEVICE

ChoicesChoices

OUTPUT DEVICEOUTPUT DEVICE

ChoicesChoices

Geometric MechanismDegrees of Freedom

Mechanism GeometricDegrees of Freedom

HUMAN INPUTMOTION, u

DESIREDOUTPUTMOTION

y(1, 2,..., n)

DATA ACQUISITION,MODELING

u(1, 2,..., n) MODELREDUCTION

u()

INPUTSUBSYSTEM

TRANSFORMATION()

EFFECTORSUBSYSTEM

1(), 2( ),..., n( )

Software: Pro/Engineer. Parametric definition of parts. Detailed geometric design capability. Part vs assembly modeling. Interfaces to analysis, FEM and CAM

packages.

2. DEVICE DESIGN AND OPTIMIZATION: Module

Synthesis of feasible candidate designs to assist the designer in selecting suitable design. Optimization and customization of the mechanism occurs here and then propogated on to the visual interface.

CAD MODULE

DESIGN MODULE

Obtain kinematic model of movement and determine appropriate input motion.

Choose appropriate output motions.

Preliminary design: select candidate mechanism.

Use virtual models to investigate the mechanism.

Customize the mechanism to the individual user and build a virtual prototype.

After testing and evaluation, build the physical prototype.

2. DEVICE DESIGN AND OPTIMIZATION: Process

a0

a1

a2

0

1

2

3

x

y

x

y HEAD BAND

r1

r2

r0

Virtual prototype ofeffector subsystem

Synthetic human model

Virtual prototype ofcoupling subsystem

Virtual prototype ofinput subsystem

Customer

Customer

VIRTUAL

WORLD

PHYSICAL

WORLD

DESIGN AND VIRTUALPROTOTYPING

INTERMEDIATEPROTOTYPE

FINAL PROTOTYPE

Designer

Physical prototype ofinput subsystem

Electronic interface

Communication

Virtual prototype ofeffector subsystem

Virtual prototype ofcoupling subsystem

Physical prototype ofinput subsystem

Physical prototype ofeffector subsystem

Physical prototype ofcoupling subsystem

3. VIRTUAL PROTOTYPING AND EVALUATION

Parametric mapping

Kinematic and dynamic evaluation with a model of the user

Design optimization and customization

Virtual and physical prototypes of the input subsystem and effector subsystem

DESIGN ENVIRONMENT

CENTRAL INTERFACE

GEOMETRIC MODELING

Cyberwarerange

scanner

Videocameras

Manipulandum

KINEMATIC MODELING

DYNAMIC ANALYSIS

CAD/CAM ProEngineer

VISUALIZATION (JACK, GeomView,

ProEnginer)

DESIGN OPTIMIZATION

Surface mesh

Motion

3-D model

simulation

MANUFACTURING

Customized components

Off-the-shelf components

Unified framework for analyzing data (geometry, kinematics, dynamics), testing (simulating) and evaluating products.

Graphical, user-friendly interface.

Heterogeneous data.

Modular.

Uses standard packages/formats.

DESIGN ENVIRONMENT: Central Interface

DESIGN ENVIRONMENT: Central Interface

CASE 1: Head-controlled Feeding Device

DESIGN SELECTION

VIRTUAL PROTOTYPING

VIRTUAL PROTOTYPING

FABRICATED PROTOTYPE

CASE 2: Head-controlled Painting Tool

DESIGN SELECTION

VIRTUAL PROTOTYPING

VIRTUAL PROTOTYPING

FABRICATED PROTOTYPE

FABRICATED PROTOTYPE

SUMMARY

Key Ideas Integrated design environment aids the designer in the rapid

realization of “one-of-a-kind” products customized to individual users

Only feasible designs are created by design module effectively reducing the optimization search space.

Virtual prototyping enables rapid evaluation within these feasible design choices.

Customized design methodology applicable to many classes of human-worn devices which need to be customized to individuals.

Limitation The component technologies are often specific to the product