Proceedings of the 2018 Ergo-X Symposium. Exoskeletons in the Workplace — Assessing Safety, Usability, and Productivity

Proceedings of the

2018 Ergo-X Symposium

Exoskeletons in the Workplace — Assessing

Safety, Usability, and Productivity

October 1, 2018 — Philadelphia, Pennsylvania

ergo

Cover Photo by The Boeing Company

Proceedings of the 2018 Ergo-X Symposium:

Exoskeletons in the Workplace — Assessing Safety,

Usability, and Productivity

October 1, 2018 — Philadelphia, Pennsylvania

Ergo-X Symposium Chairs

Christopher Reid, e Boeing Company

David Rempel, University of California, Berkeley

Kermit Davis, University of Cincinnati

Proceedings Editors

Brian Lowe, National Institute for Occupational Safety and Health

William Billotte, National Institute of Standards and Technology

George Brogmus, Liberty Mutual Insurance

omas McDowell, National Institute for Occupational Safety and Health

Christopher Reid, e Boeing Company

David Rempel, University of California, Berkeley

Divya Srinivasan, Virginia Tech

DEPARTMENT OF HEALTH AND HUMAN SERVICES

Centers for Disease Control and Prevention

National Institute for Occupational Safety and Health

Proceedings of the 2018 Ergo-X Symposium

This document is in the public domain and may be freely copied or

reprinted.

Disclaimer

ese proceedings do not constitute endorsement of the views expressed or recom-

mendations by the National Institute for Occupational Safety and Health (NIOSH),

Centers for Disease Control and Prevention (CDC)or the Human Factors and Ergo-

nomics Society (HFES). e opinions and conclusions expressed in the presentations

are those of the presenters and not necessarily those of NIOSH or HFES. Mention of

any specic company name or product does not constitute endorsement by NIOSH or

HFES. Summary content for the sessions was developed by the Editors and where con-

tent is attributed to a specic presenter the presenter was provided the opportunity to

review and edit the content. Recommendations are not statements of NIOSH or HFES

policy, or of any agency or individual involved. e information is intended to be used

in advancing the knowledge needed for improving the safety and health of workers

using exoskeletons.

Get More Information

Find NIOSH products and get answers to workplace safety and health questions:

1-800-CDC-INFO (1-800-232-4636) | TTY: 1-888-232-6348

CDC/NIOSH INFO: cdc.gov/info | cdc.gov/niosh

Monthly NIOSH eNews: cdc.gov/niosh/eNews

Suggested Citation

NIOSH [2019] Proceedings of the 2018 Ergo-X Symposium: Exoskeletons in the

Workplace — Assessing Safety, Usability, and Productivity. By Lowe B, Billotte W,

Brogmus G, McDowell T, Reid C, Rempel D, Srinivasan D (Editors). Cincinnati, OH:

U.S. Department of Health and Human Services, Centers for Disease Control and

Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH)

Publication No. 2020-102, https://doi.org/10.26616/NIOSHPUB2020102.

DHHS (NIOSH) Publication No. 2020-102

DOI: https://doi.org/10.26616/NIOSHPUB2019102

October 2019

iiiProceedings of the 2018 Ergo-X Symposium

Preface

Exoskeletons and Exosuits are wearable technologies designed to augment the hu-

man musculoskeletal system to improve physical performance. eir potential value

extends across disciplines, and includes improving industrial worker capabilities and

enhancing medical rehabilitation. e concept is not new; patent designs are traceable

to the 1800s. but they moved from science ction to reality in the 1970s with research

and development by the military to enhance warghter capabilities and have taken

o in the past decade in the industrial sector with improved designs and materials.

Now many manufacturers are investing heavily in the evaluation of exoskeletons for

assembly and warehouse jobs to reduce fatigue and injury and improve productivity.

In response to the rapid market growth, industrial users and researchers unied in

late 2017 to form the ASTM F48 Exoskeletons and Exosuits Standards Committee to

create guidelines on the safe design and adoption of this wearable technology.

For the past 3 years, Boeing, Ford, Toyota, and BMW on the private industry side,

Navy and Army on the military side, and a number of medical institutions, have been

conducting research on exoskeleton and exosuit technology. is work, including

user assessment and implementation ndings has been primarily internalized to each

institution, though some of their broader ndings have been shared publicly. During

discussions in early 2018 with ASTM F48 and the Human Factors & Ergonomics

Society (HFES) executive leaderships, it became apparent that there was an oppor-

tunity for both organizations and their members to leverage each other’s expertise to

help accelerate exoskeleton and exosuit design and standards using human factors and

ergonomics principles in user-centered design. ese principles were already being

actively explored in the ASTM F48.02 subcommittee task groups on exoskeleton and

exosuit anthropometric size and shape, usability, ergonomics, safety, and training.

is joint coordination and planning led to the rst national symposium, titled “ErgoX

Symposium: Exoskeletons in the Workplace – Assessing Safety, Usability, & Produc-

tivity,” which was held on October 1, 2018 prior to the Human Factors & Ergonomics

Annual Meeting in Philadelphia, PA. e ErgoX symposium provided a forum for

designers, users and researchers to share insights and ndings in a public setting on

human factors issues related to exoskeleton and exosuit technology. e format in-

cluded TEDx style presentations with discussion panels and product interaction with

developers and vendors in the same oor space. Experts from the military, medical,

and industrial domains participated as users, developers, regulatory, or university

sector speakers and attendees.

e content of the symposium attracted support and participation from many com-

panies (e.g. Liberty Mutual, Boeing, Mawashi, Levitate), U.S. Federal agencies, (e.g.

NIST, FDA, NIOSH, VA, etc.), universities, international Federal agencies, and sci-

entic bodies (National Academy of Sciences). With a little over 130 attendees, this

initial symposium included both US (84%) and non-US (16%) attendees, with from the

latter comprising eight countries in Europe, Asia, and the Caribbean. e majority of

attendees were from private industry (34%), followed by academia (29%), government

(27%), and then exoskeleton developers (8%). Based on exoskeleton domain of inter-

est, attendees and their organizations were grouped into the industrial (61%), military

(27%), and medical (10%) representations.

Proceedings of the 2018 Ergo-X Symposium

is report summarizes important points made by presenters and panelists. the suc-

cess of the exoskeleton and exosuit symposium and the rapid rate of development of

this technology and research has led the organizers to plan another Symposium prior

to the next HFES meeting in Seattle, Washington on October 28, 2019, continuing in

subsequent years. Lessons learned from this year will carry forward into the program

for 2019, with the intention to continue to demonstrate iterative progress for user-cen-

tered design and industry standards.

Christopher R. Reid, PhD

Secretary-Treasurer, Human Factors and

Ergonomics Society, Symposium Chair

David Rempel, MD

Symposium Co-Chair

Kermit Davis, PhD

President, Human Factors and Ergonomics

Society, Symposium Co-Chair

vProceedings of the 2018 Ergo-X Symposium

Executive Summary

e Proceedings of the 2018 Ergo-X Symposium: Exoskeletons in the Workplace have

been assembled to disseminate the speakers’ presentations and to summarize the

question and answer/discussion periods that followed the presentations within each

session. e proceedings appear by session and include summary points with links to

presentation slides from speakers who agreed to provide them. e Ergo-X Proceed-

ings Editors identied and documented the summary points and gave presenters of

specic content (such as keynote presentations) an opportunity to review, edit, and

approve the content.

Here are some of the key summary points from the 2018 Ergo-X Symposium:

■ Metabolic demand may be a predictor of fatigue onset; however, we need a bet-

ter understanding of how the positive or negative eect of an exoskeleton on

metabolic demand aects injury prevention/risk.

■ e t of the exoskeleton system is complex. Static assessments of t that do not

consider task dynamics are insucient; multivariate anthropometric data are

critical to t.

■ Simulation and digital human modeling technologies have potential use in (1)

assessing the interface between the user and exoskeleton and (2) reducing the

test and evaluation burden of using human subjects.

■ Existing exoskeleton systems require a period of adaptation by the end user. For

a new user, task performance is not likely to reach a steady state immediately. We

need to establish acceptable test durations for exoskeleton trials.

■ Cognitive and psychomotor eects of exoskeleton use have been observed and

are likely task dependent.

■ Industrial exoskeleton designs should be compatible with o-the-shelf tools,

equipment, and personal protective equipment, rather than relying on specialty

tools and custom interfaces.

■ Although industry speakers presented examples of wider-scale deployment of

overhead support exoskeletons, overhead work with tool support appears to be

the most mature industrial-use case at present.

■ e FDA oversees devices marketed/prescribed for medical use. Early adoption

of medical exoskeletons may be more promising among individuals who are less

adapted to other mobility-assistive technologies for their disabilities.

■ In the rehabilitation domain, clinics can utilize exoskeletons to assist therapists

in delivering appropriate therapeutic doses.

■ ASTM Committee F48 on Exoskeletons and Exosuits and other standards

organizations oer a forum for sharing exoskeleton knowledge.

Proceedings of the 2018 Ergo-X Symposium

Feedback gathered from attendees and participants revealed 19 dierent topics (see

the word cloud) that were issues or concerns for exoskeleton developers, researchers,

and end users in 2018 and moving forward. e top four topics were (1) return on

investment (ROI) considerations; (2) size, shape, and t of exoskeletons on users; (3)

longitudinal eects of exoskeleton usage; and (4) “What metrics are right?” for mea-

suring safe, eective, or reliable system design and integration for users or patients.

Technology needs, gaps, and concerns of the 2018 Ergo-X Symposium participants.

viiProceedings of the 2018 Ergo-X Symposium

Table of Contents

Preface .................................................................. iii

Executive Summary ....................................................... v

Acknowledgments ........................................................ viii

List of Speakers by Session ................................................. 1

Conference Sessions ....................................................... 3

Keynote 1, Wearable Robotic Systems: Global Landscape and Opportunities . 3

Opening, Update on ASTM Committee F48 on Exoskeletons and Exosuits

and other Standards Eorts

............................................ 5

Keynote 2, Ergonomic Assessment of a Space Suit: From the Perspective of

Population Analysis, Fit, Accommodation, Comfort, and Performance

...... 6

Exoskeleton User Discussion Panel ..................................... 8

Research Methods 1—Design for Population Accommodation & Performance 13

Exoskeleton Developer Discussion Panel ................................ 16

Research Methods 2—Assessing System Usability ........................ 18

Research Methods 3—Assessing Safety Panel ............................ 23

Research Methods 4—Assessing Ergonomics ............................ 25

Closing Discussion Panel .............................................. 28

viii

Proceedings of the 2018 Ergo-X Symposium

Acknowledgments

Important contributors to the development of this document were Lois Smith and

Julie Freeman (former HFES administration); Steve Kemp, HFES; and Dawn Castillo,

Hongwei Hsiao, Kathleen Goedel, Seleen Collins, Elizabeth Clements, and

Vanessa Williams, NIOSH.

Ergo-X Symposium sponsorship was provided by the

following:

Liberty Mutual Insurance

e Boeing Company

Mawashi Science & Technology

Levitate Technologies, Inc.

ASTM International

Northern Illinois University

National Academy of Sciences, Board on Human Systems Integration

1Proceedings of the 2018 Ergo-X Symposium

List of Speakers by Session

Session Speakers

Keynote 1 (morning) Bruce Floersheim, GoXStudio

Opening Session William Billotte, NIST

Keynote 2 (afternoon) Sudhakar Rajulu, NASA Johnson Space Center

Exoskeleton User

Discussion Panel

Kendra Betz, U.S. Veterans Aairs

Robert Schram, Toyota

Ron Zmijewski, U.S. Navy Human Assistive

Technology

Moderator: Robert Fox, General Motors

Research Methods 1—

Design for Population

Accommodation &

Performance

Monica Jones, University of Michigan

Joseph Parham, U.S. Army Natick Soldier Research

Development & Engineering Center (NSRDEC)

Lei Stirling, Massachusetts Institute of Technology

Moderator: Krystyna Gielo-Perczak, University of

Connecticut

Exoskeleton Developer

Discussion Panel

Brandon Frees, Ekso Bionics

Chris Beaufait, Sarcos Robotics

Marty Linn, General Motors

Ignacio Galiana, Harvard Wyss Institute

Moderator: Christopher Reid, e Boeing Company

Research Methods

2—Assessing System

Usability

Alix Dorfman, Underwriters Lab (UL), Wiklund

Kadon Kyte, e Boeing Company

Kevin Purcell, U.S. Army Public Health Center

Moderator: Carisa Harris-Adamson, University of

California, San Francisco/University of California,

Berkeley

Proceedings of the 2018 Ergo-X Symposium

Session Speakers

Research Methods 3—

Assessing Safety

Ian Marcus, U.S. Food & Drug Administration

Roger Bostelman, NIST

Angela Boynton, U.S. Army Research Laboratory

Moderator: Brian Lowe, NIOSH

Research Methods 4—

Assessing Ergonomics

Marty Smets, Ford Motor Company

Maury Nussbaum, Virginia Tech

Karen Nolan, Kessler Foundation

Moderator: Cathy White, Dow Chemical Company

Closing Discussion Panel Delia Treaster, Ohio Bureau of Workers Compensation

Gerard Francisco, TIRR Memorial Hermann Hospital

Donald Peterson, Northern Illinois University/ASTM

Committee F48 on Exoskeletons and Exosuits

Moderator: Cindy Whitehead, U.S. Navy–Naval Sea

System Command

Editors’ Note: e symposium organizers invited all speakers to submit their presentation content and to include

this content in session summaries. e presentation slides and summaries included here are for speakers who gave

permission to include their content.

3Proceedings of the 2018 Ergo-X Symposium

Conference Sessions

Session Title

Keynote 1, Wearable Robotic Systems: Global Landscape and Opportunities

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Keynote_Bruce_Floersheim.pdf

Speaker

Bruce Floersheim, WearRAcon/GoXStudiom

Summary Points

■ A show of hands indicated the audience was a relatively equal mix of govern-

ment, industry, and academia.

■ e “wow factor” of exoskeleton technology helps from a marketing standpoint.

In addition, the movie industry has increased the technology’s visibility.

■ Drivers of the industry include public curiosity and the fact that people are living

longer. As their bodies break down, they still want to be able to do the same

activities, and therefore the technology has advanced.

■ North America is playing a bit of a catch-up game in comparison with the rest

of the world in this arena. Europeans and Asians have been doing this work on a

scale that is more organized and integrated.

■ Insurance costs are increasing across the board. Insurance companies are

showing more interest in these technologies.

■ Many labs are focusing on “return function,” that is, trying to return some

function to users such as stroke survivors and patients with paraplegia.

■ is goal is common for lower-body systems.

■ An “enhance function” focus puts emphasis on fully functioning users and taking

them to increased capability.

■ e desire for enhanced quality of life is slowly trickling into this technology.

■ Getting access to the technology will be a factor in consumer adoption; industrial

users will be the rst and main adopters.

■ Large companies are procuring systems and testing them in warehousing and

manufacturing facilities.

Proceedings of the 2018 Ergo-X Symposium

■ Computing feedback and controlling the feedback loop are important factors for

fully optimizing and individualizing the technology for particular wearers.

■ At present, the power source is probably the Achilles heel (weak point).

■ We need to gure out better ways to provide power in order to shrink the size

and cost of systems.

■ Industry is leading the drive to promote development and adoption of these

systems.

■ One goal is to allow an aging workforce to continue to do physically demanding

jobs, as long as they desire to continue such work.

■ Exo-assist technologies are of interest to labor organizations as a means of im-

proving worker capability and quality of life, without fully replacing the worker.

■ Question: Can these technologies broaden the demographics for persons who

are capable of performing highly demanding jobs?

■ Stakeholders should seek involvement with standards development (ISO,

ASTM).

■ Traditional industrial robotics manufacturers may start making their presence

known in this technology area.

■ ere are many start-ups in the commercial market in the United States and

Europe, where there is more access to capital investment. In Asia, technologies

are primarily coming out of “old line” industrial companies.

■ e companies are developing the technologies for their own workforce but are

also starting to look for external sales opportunities.

■ Exo-system technologies are out there. Question: What can be done to make

them “seamless”?

Q&A/Discussion

■ Question: Does a decreased metabolic rate correlate with or conrm a reduction

in injury? We are still at the early end of understanding this relationship, and a

number of pilot programs are looking at this. Fatigue is an indicator of an in-

creased risk of injury. We are operating on the assumption that reducing met-

abolic cost increases the time to get to the fatigued state and therefore reduces

injury. We need more data to assess rates of injury.

5Proceedings of the 2018 Ergo-X Symposium

Session Title

Opening, Update on ASTM Committee F48 on Exoskeletons and Exosuits

and other Standards Efforts

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Keynote_Bill_Billotte.pdf

Speaker

William Billotte, NIST/ASTM Committee F48 on Exoskeletons and Exosuits

Summary Points

■ “We are the future” platform.

■ Exoskeletons will use quantum computing, articial intelligence, and

high-performance computing.

■ We need standards to ensure exoskeletons are safe, are reliable, and perform

as intended.

■ e ASTM Committee F48 on Exoskeletons and Exosuits was established in

September 2017.

■ F48 subcommittees were formed around a life-cycle approach. e speaker gave

an overview of each subcommittee.

Q&A/Discussion

■ Terminology work is ongoing, such as dening an exoskeleton and determining

whether an exosuit is a type of exoskeleton.

■ Participants discussed the scope of ASTM F48. It likely excludes traditional

prosthetics, but new prosthetics that are more akin to wearable robotics would

fall under its scope.

■ Standards are meant to facilitate innovation. We need to investigate what

metrics are going to work best for the exoskeleton community.

■ Standards are living documents to revise and improve over time.

Proceedings of the 2018 Ergo-X Symposium

Session Title

Keynote 2, Ergonomic Assessment of a Space Suit: From the Perspective of

Population Analysis, Fit, Accommodation, Comfort, and Performance

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Keynote_Sudhakar_Rajulu.pdf

Speaker

Sudhakar Rajulu, NASA Johnson Space Center

Summary Points

■ A spacesuit is a type of exosuit. NASA has been dealing with an ensemble system

in the spacesuit for many years.

■ e t of the suit in a static situation may not be the same as in a dynamic

situation.

■ We need to make donning and dong simpler.

■ In microgravity, the legs need little mobility. e arms—shoulders, elbows, and

wrists—need the most mobility. Joint bearings will be necessary for lower-limb

mobility for the Mars expedition.

■ e Russian Space Agency selects cosmonauts who have very similar, restrictive

anthropometric proles.

■ e Apollo program used a similar restrictive sizing approach, combined with

minor adjustments/alterations. Since the Shuttle program, NASA has t a wider

range (5th percentile female vs. 95th percentile male). NASA made its equip-

ment sizing inclusive of a wider anthropometric variation.

■ Each spacesuit costs several million dollars.

■ e Shuttle suit incorporated a Hard Upper Torso (HUT) instead of a soft suit

upper component.

■ For the Mars mission, the suits need to be modularized and highly adjustable

so that individual suit components can be interchanged and exchanged. It is

dicult to carry spare parts for individual crewmembers because of payload

considerations.

■ e current approach to studying and improving suit t issues is merging full

body scans with CAD models of the hard suit components.

■ Shoulder topography changes with movement from static to dynamic poses.

■ Models under development will simulate what happens to the shoulder as the

user goes through a dynamic motion. We will be able to apply this to the entire

population.

7Proceedings of the 2018 Ergo-X Symposium

■ Monte Carlo simulations represent variation across the entire body size and

shape range for a particular suit type, to determine who will t into it.

■ When designing a suit or exoskeleton, we need to understand how it restricts or

limits natural movement.

■ e goal is for an exoskeleton to t everyone, not a specic subpopulation.

■ Because testing everyone is too time consuming, we need to improve methods of

virtual simulation with the entire population.

Q&A/Discussion

■ Use of the hands is important in suited operations. Augmenting hand function

is a priority. Upper arm (shoulder) excursions and mobility are also under study.

e NASA exercise countermeasures group has considered looking at exoskele-

tons for rehabilitation purposes to counteract the muscle atrophy and bone loss

expected in a journey to Mars.

■ What should the next steps be for exoskeletons? We need to enable users to don

and do systems without assistance, and we need to make sure the active modes

don’t cause problems.

Proceedings of the 2018 Ergo-X Symposium

Session Title

Exoskeleton User Discussion Panel

Moderator

Robert R. (Bob) Fox, General Motors

Speakers

Robbie Schram, Toyota

Introducing Exoskeletons into the Toyota Manufacturing Environment

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/User_Discussion_Robert_

Schram.pdf

Kendra Betz, U.S. Veterans Affairs

Exoskeletons as Assistive Technology for Rehabilitation: Clinical Perspectives

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/User_Discussion_Kendra_Betz.pdf

Summary Points

■ In 2016, Toyota built 2 million vehicles and sold 2.5 million, involving 40,000

workers across 13 manufacturing locations.

■ e focus of Toyota exoskeletons has been on the shoulder, upper back, and arm.

■ Toyota has used passive upper-body exoskeletons for overhead work and for

under-chassis work.

■ e company conducted trials of exoskeletons in seven plants, with the most

deployment in Canada, Indiana, Kentucky, and Texas.

■ It has deployed 239 exoskeletons in the United States and Canada, which are in

use now in Paint and Assembly; it will deploy up to 440 in 2019.

■ Nearly all deployed exoskeletons are the Levitate device; an additional 15

Eksobionics Eksovests have also been deployed.

■ e Levitate AirFrame trial

— Toyota has tested the Levitate AirFrame in 10 processes, of which more

than half involved “no good” shoulder postures.

— e company used subjective team member feedback (rating perceived

exertion, discomfort, and equipment usefulness) at three of its motor

manufacturing plants (in Kentucky, Indiana, and Canada).

9Proceedings of the 2018 Ergo-X Symposium

— Results showed that 70% to 80% of team members preferred using the

device. ey also showed a 73% average reduction in perceived exertion

and 44% reduction in discomfort ratings.

— Iowa State also collected %MVC (maximum voluntary contraction)

measurements via electromyography (EMG) with use of the AirFrame at

the plant in Canada.

— EMG results showed that with the AirFrame an overall reduction in

%MVC (that is, decreased fatigue) occurred for the shoulder (5.8%) and

the back (4.1%) but not for the biceps (for which %MVC increased for

certain tasks).

— Strength tests showed no reductions and some increases over the course

of the study.

■ Toyota and Levitate co-developed “mutilation covers” to reduce damage (from

contact between the exoskeleton and the product—cars) and to increase user

condence.

■ A second EMG trial used %MVC, Toyota’s internal ergo assessment tool (TEBA),

and ACGIH’s Upper Limb Localized Fatigue equation.

— Researchers mapped TEBA to ACGIH’s Upper Limb Localized Fatigue

TLV (reshold Limit Value) in order to use TEBA to screen jobs that may

benet from exoskeleton use.

■ e plan for moving forward involves these actions:

— Inventory all jobs using TEBA criteria or injury data and calculate TLV.

— Use the Hierarchy of Controls before using an exoskeleton (which

Toyota considers a form of PPE); perform a risk assessment to ensure no

new hazards emerge with the exoskeleton.

— Finalize standards (mandatory or voluntary); monitor long-term out-

comes; put dierent types/models to trial.

■ Exoskeletons are a form of assistive technology for use by individuals with

disabilities.

■ For these applications we use the HAAT model: a Human performing an Activity

with an Assistive Technology within a context.

■ Candidates for exoskeletons are often full-time wheelchair users, with little or

no ability to walk, or those who are regaining the ability to walk through

rehabilitation.

■ e FDA has cleared four exoskeleton devices: ReWalk, Indego, Ekso, and

Hybrid Assistive Limb (HAL) by Cyberdyne. HAL, the device most recently

cleared, requires an overhead harness support system and therefore is for

indoor use.

■ Clinical decisions incorporate research, professional experience, and especially

the client experience and perspective.

Proceedings of the 2018 Ergo-X Symposium

■ A video showed the process for putting the exoskeleton on and the necessity of

assistance for the user to go from sitting to standing and walking.

■ Exoskeletons are useful for a subset of individuals with disabilities; they must

meet certain criteria. Individuals who have had more time to adapt to their

disability are more likely to be okay with use of a wheelchair but are interested in

advances in exoskeleton technology.

■ In some cases, there is a steep learning curve for clinicians and clients because of

the devices’ complexity.

■ It is critical for all device options to be available for trial, to match the candidate

to the optimal exoskeleton technology.

■ Current exoskeletons do not yet support the client to move at normal walking

speeds, but this feature will improve as the technology advances.

■ Exoskeleton users need a trained companion, a limitation for deployment to

certain environments and circumstances.

■ Skin injury and protection are signicant concerns for clients with neurologic

injuries.

■ e VA has developed the VA National Clinical Protocol, with resources to help

clinicians with evaluation, training, and inclusion/exclusion criteria.

■ Being able to lease the exoskeleton for an extended trial before purchasing it has

been extremely helpful for successful implementation.

■ Many exoskeleton research projects continue, including a randomized control

trial (RCT) at 10 VA sites led by Drs. Spungen and Asselin.

■ Here are a few other points about exoskeletons and ethics:

— Better technology is always on the horizon; limited competition = high

costs (in addition to care, repair, and maintenance costs).

— Consider the “research while implementing” ethics of developing stan-

dards while implementation is taking place; we need responsible, experi-

enced participants in the standards-making process.

— CLOUT (Clinical Limits of Use Tools) provides a matrix of application

functions/limitations for each device.

Q&A/Discussion

■ Here are some common threads among exoskeleton users:

— Users must like the device to use it and benet from it.

— End users and vendor/manufacturers must share feedback with each other.

— e complexity of the workplace aects user adoption.

— Embrace the technology and work to advance its adoption to increase its

eectiveness.

11Proceedings of the 2018 Ergo-X Symposium

■ Advice to vendors/manufacturers included the following:

— Listen to users about features, t, and design details: keep your ear to the

ground.

— Interact with users in the worksite.

— Tailor devices to applications.

— Intuitive usability is currently a complexity and challenge to overcome.

■ Question: What is the minimum exposure the user will need to have with the

device to provide useful feedback?

— Toyota uses a slow ramp-up, from 30 minutes to a full 2 hours and up to a

month for industrial use, and from a few days to a week for military.

— In the medical domain, a functional minimum for “lease-to-own” users is

that they must be able to get in the device with minimal assistance, achieve

a sit-to-stand, and walk 10 meters with minimal supervision.

■ Question: How do you manage the program you have? And do you expect that

ramping up the same program will support it?

— Internal standards provide guidelines on the lifecycle of the program.

— A selection process helps identify potential users.

— On the production oor, we have point people at the main plants, along

with medical, safety, and shop reps.

— We are already out of the testing phase and are in the implementation

phase, so we are expecting this model to t.

■ Question: In terms of outcomes, how do exoskeletons compare to traditional

physical therapy?

— Outcomes are dierent from those with traditional therapy and vary more.

Users have increased capabilities but not at a functional norm, so some

prefer the mobility they have achieved in their wheelchair.

■ Question: How are you selling these devices to leadership? In the medical do-

main, how can you make the economics more feasible?

— Toyota has added a digital transformation and human mobility “pillar” to

the company; it is not just a car company but also a mobility provider. is

is a cultural priority, so it is now easier to get a “foot in the door.” e exo-

skeletons continue to provide ROI and are getting better. ere is a bit of

competition; when leadership sees a competitor’s video on their exoskele-

ton deployment, it is more likely to support adoption.

— In the medical domain, we need more competition to drive innovation. We

have too few models to choose from, and this is driving up costs.

Proceedings of the 2018 Ergo-X Symposium

■ Question: Do you think you aect the lower extremities and change

posture with an exoskeleton?

— We did not do any lower-extremity evaluation. We are looking at strength

testing and injuries. Short-term, we have not seen any strength decre-

ments, but long-term evaluation will be necessary.

■ Question: Do you actually see user dropout? If so, why?

— Toyota: Yes, some don’t want to use devices, but others use them full-time.

Fit and comfort are the usual reasons for rejection, and we wonder about

whether we should make device use mandatory.

13Proceedings of the 2018 Ergo-X Symposium

Session Title

Research Methods 1—Design for Population Accommodation & Performance

Moderator

Krystyna Gielo-Perczak, University of Connecticut

Speakers

Monica Jones, University of Michigan

Three-Dimensional Anthropometric Data for Exoskeleton and Exosuit Design

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Population_Accommodation_

Monica_Jones.pdf

Joseph Parham, U.S. Army Natick Soldier Research Development &

Engineering Center (NSRDEC)

Anthropometric Considerations in Exoskeleton Development

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Population_Accommodation_

Joe_Parham.pdf

Leia Stirling, MIT

Quantifying Physical and Cognitive Fit for Assessing Exoskeletons

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Population_Accommodation_

Leia_Stirling.pdf

Summary Points

■ Human measurement and modeling methods are an opportunity to inform the

design and evaluation of exoskeletons.

■ No two people are the same size and shape.

■ e Anthropometric Survey of U.S. Army Personnel (ANSUR) and other data

sets are typically one-dimensional.

■ ree-dimensional (3D) scans give richness of size and shape.

■ Models and data are publicly available at www.humanshape.org. ese data can

be exported for 3D shape in computer-aided design (CAD) systems.

■ Range-of-motion scans (dynamic) show what happens to the shape when the

joint moves through the range of motion.

Proceedings of the 2018 Ergo-X Symposium

■ No actual person looks like a 5-percentile or 95-percentile body model. ere

are approaches for morphing body shape models from GHBMC (Global Human

Body Models Consortium) and Total Human Model for Safety (THUMS) onto

nite element models so they are more representative of a broader population.

■ We need dynamic, rapid-loading models for crash tests. We can use quasi-stat-

ic loading models for exoskeletons at a mechanical interface to quantify shape

deformation, esh deformation, and pressure points through nite element (FE)

modeling.

■ A skeletal geometry model with CT and MRI produces a parametric representa-

tion of the skeletal system as a function of age, gender, and other variables.

■ ese models and data provide opportunities to develop quantitative measures

of t, to customize t, and, at a population level, to inform sizing requirements.

■ ANSUR is one of the largest anthropometric survey databases in the world. e

survey was conducted in 2011–2012 and collected data on 12,000 soldiers, in-

cluding 93 anthropometric measurements and 3D scans.

■ Typical exoskeleton testing involves small samples. We can reference anthropo-

metric databases to see where those subjects fall within a population.

■ Users of equivalent height and weight can vary substantially across other anthro-

pometric variables.

■ Common measurements obtained in a semi-nude condition are not reective of

the full gear and equipment required of a soldier. What is the delta between the

two conditions? We refer to this as “encumbered anthropometry.”

■ Dynamic t and static t are dierent in terms of joint alignment with the

system. Statically determined alignment may not predict alignment in dynamic

tasks.

■ Cognitive t is a consideration: Does use of the system aect ability to perform

other psychomotor tasks?

■ Users vary in the time they require to learn how to use systems. e design of

active controllers can be based on the individual user’s style of adaptation.

■ Spacesuit t is based on observations of the suit technician and subjective

feedback from the suit wearer. Methods have been developed to provide the suit

technician with quantitative information about what is not visible from outside

the suit. Technicians are interested in relative motion between the human and

the encasing suit.

■ Relative coordination measurements involve collecting inertial measurement

unit (IMU) data on the human and IMU data on the suit to evaluate relative

motion between the two.

■ An example from the gait cycle is at the point of heel o—hypothesized to be the

foot lifting out of the boot as the suited subject is walking. ese measurements

align with subjective response of the suit user.

15Proceedings of the 2018 Ergo-X Symposium

■ An open question: How does t aect performance?

■ Exoskeletons may aect visual attention and reaction time performance of some

users, with wide variability. Some tests of exoskeleton use have shown an impact

on inherent cognitive capabilities.

■ Perceived increase in overall workload has been assessed with the NASA TLX

scale.

Q&A/Discussion

■ ere are a number of population-level descriptions for anthropometry, but we

also need population-level descriptions of biomechanics.

■ In a short-term (3-day) study, users may not achieve adaptation or steady-state

performance with the system, but the study may indicate immediate “intuitive-

ness” of use. e underlying t to the person aects the rate of adaptation, or of

achieving steady-state performance.

■ Customers ask vendors about the impact of exoskeleton use on quality of work.

is is likely task-specic. Ideal t may depend on the task. Dierent operational

performances may require dierent adjustments or alignments.

■ Question: What are specic variables to measure for addressing adaptation to

exoskeleton systems? Individuals have dierent responses to perturbations,

and this variability may be important to the design of control systems. ere are

examples of humans responding as both overdamped or underdamped systems

when subjected to perturbation.

Proceedings of the 2018 Ergo-X Symposium

Session Title

Exoskeleton Developer Discussion Panel

Moderator

Christopher R. Reid, The Boeing Company

Speakers

Chris Beaufait, Sarcos Robotics

Achieving Technical and Manufacturing Readiness for the

Commercialization of Powered Exoskeletons

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Developer_Discussion_Chris_

Beaufait.pdf

Marty Linn, General Motors

Roboglove—A Human Grasp-Assist Device

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Developer_Discussion_Mar-

ty_Linn.pdf

Brandon Frees, Ekso Bionics

Exoskeletons—Ideas for Implementation and Change Management

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Developer_Discussion_Bran-

don_Frees.pdf

Summary Points

■ Developers of powered exoskeletons aim to combine human strong points such

as instinct, intelligence, and judgment with robotic attributes such as strength,

endurance, and precision.

■ Developers want to augment worker performance while keeping workers out of

harm’s way.

■ A major goal is to reduce occupational injuries and musculoskeletal disorders.

One way to achieve this is to reduce worker fatigue and fatigue-related injuries.

■ Developers also want to develop exoskeletons that can decrease times for users

to recover from injuries and musculoskeletal disorders.

17Proceedings of the 2018 Ergo-X Symposium

■ One of the challenges of powered exoskeleton development is to increase

eciency in order to reduce power consumption. is will allow the use of

completely untethered units and the use of smaller, lighter battery packs.

■ Powered exoskeleton designs should ensure that the exoskeleton follows the

movements of the worker and not vice-versa.

■ Exoskeletons need to be adaptable to diverse environments.

■ Designs should be compatible with o-the-shelf tools and equipment; avoid spe-

cialty tools and custom interfaces when practical.

■ Another goal is to increase worker productivity.

■ Designs should allow the worker to operate at a normal pace and within the

existing work environment.

■ Designs should limit modications to normal work operations; exoskeleton use

should be integrated with existing work processes instead of forcing employers

to alter work operations to accommodate exoskeleton use.

■ One of the major challenges for developers is to create worker buy-in.

Q&A/Discussion

■ Eorts are underway to integrate wearable sensors into exoskeletons. Developers

and researchers are collecting data on muscle activity, motion, force, metabolic

metrics, and so on. New, exible electronics will facilitate such exoskeleton de-

signs. e use of instrumented under-suits is also being explored.

■ Exoskeleton designs and applications must evolve to ensure user buy-in.

■ Exoskeleton developers hope that exoskeletons will eventually become

mandatory PPE in certain work processes and environments.

■ Exoskeleton developers must create pathways to move designs from the research

and development realm into the user space. ere must be “pay points” that

demonstrate worker benets. Developers must identify and dene problems,

and exoskeleton use must yield solutions to those problems. At the same time,

exoskeletons must have realistic price points.

■ Developers must address psychosocial aspects of exoskeleton use. Interactions

among exoskeleton users and the social environment are important and should

be evaluated.

■ Exoskeleton aesthetics are important to some users. Sleek designs are typically

favored, but opinions are mixed.

Proceedings of the 2018 Ergo-X Symposium

Session Title

Research Methods 2—Assessing System Usability

Moderator

Carisa Harris-Adamson, University of California

Speakers

Kevin Purcell, U.S. Army Public Health Center

Exoskeleton Usability: Task Differences and Anthropomorphism

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Usability_Kevin_Purcell.pdf

Kadon Kyte, The Boeing Company

Perspectives on Exoskeleton Usability: Insight from Boeing Factory

Introduction

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Usability_Kadon_Kyte.pdf

Alix Dorfman, Underwriters Lab (UL), Wiklund

Assessing System Usability: Research Methods and Special Considerations

for Rehabilitative Exoskeleton Evaluation

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Usability_Alix_Dorfman.pdf

Summary Points

■ e ISO 9241-11 denition of usability emphasizes achieving “…specied goals

with eectiveness, eciency and satisfaction in a specied context.”

■ Exoskeletons have been scoring at a 4 or 5 on a 7-point Likert scale for usability.

Higher usability is critical to exoskeleton success.

■ If usability is poor, the exoskeleton will not be adopted.

■ ere is a strong connection between reduced metabolic cost and a higher

usability rating.

■ Methods to assess metabolic cost/energy expenditure include these:

— VO2

— Core temperature

19Proceedings of the 2018 Ergo-X Symposium

— Augmentation factor—developed by Dr. Mooney and colleagues;

predictive measure of power transfer to the user from the exoskeleton.

— PoLoTAE—Position and Load Test Apparatus for Exoskeletons (developed

by NIST)

■ For medical and therapeutic devices, usability should also consider the individual

assisting with the therapy.

■ In military applications, there is not much literature on usability yet.

■ Marketing classications adopted from usability practitioners include Features

vs. Benets:

— Feature—a previously specied task that an exoskeleton can help the user

perform, which tends to be related to a xed task. is contrasts with Ben-

et—what the user wants in the rst place (solution to a user’s problem),

which tends to involve a dynamic task with unspecied subtasks.

■ An application example is the original Lockheed HULC, which was designed

to increase a warghter’s ability to carry a large load while minimizing fatigue.

e HULC accomplished that specic goal. However, the soldier/warghter has

other tasks beyond going from point A to point B, such as accomplishing many

sub-goals.

■ One company’s approach is to identify where the issues are in terms of

geographical location and with respect to the individual.

— First, identify high-risk tasks by using safety data. Look at all safety in-

cidents and pinpoint by body part, various risk factors, and geographic

location to dene issues to address.

— Second, strategically align with dierent programs.

— ird, apply the best technology for the right application.

■ Industrial work types include these examples.

— Postural assist activities: maintaining a static posture for extended periods,

such as when working with overhead wire bundles, using ne motor skills,

and working near the ground.

— Hand tool usage: riveting/bucking, drilling, torqueing

— Equipment relocation/moving: workstands, tool/equipment carts

— Manual material handling: Lifting/lowering, pushing/pulling, carrying

■ Industrial use cases include these examples.

— Systems installations: doors, seats, lavatories

— Structures work: drilling, installing fasteners

— Paint work: sanding, masking, painting

— Move teams (hoping to get more into this application area): work stands,

carts, large-part moving and assembly

Proceedings of the 2018 Ergo-X Symposium

■ User-centered design goal: We need to embed the end user into the entire prod-

uct development process, to reduce injuries, maximize human performance and

product quality, and improve ease in implementation into production systems.

■ User-centered design approach: Start from the initial meeting with the supplier,

before the system is brought into the lab for testing. Have the perspectives of the

end user in mind through all aspects of the exoskeleton evaluation. When a com-

pany comes on site to demonstrate an exoskeleton system, two or three mechan-

ics will participate to give candid feedback, enabling initial down selection.

— Specication of requirements

— Design

— Simulation

— Prototype building

— Human testing

■ Usability testing approach: Involve end-users throughout.

— Site visit: supplier and end-user feedback; quick assessment

— Lab assessment: simulated environments; end-user test pool

— Field assessment: production shop trial

— Testing over time, to address problems

— Targeting workers who have a negative experience, to make them

champions

■ One speaker described results of a 6-month trial of a shoulder vest system.

■ Usability metrics: primarily subjective (Likert Scale: 1–7)

— Usefulness, adjustability, restriction, t, thermal, balance, comfort, overall

rating

— Metrics also included open-ended responses

■ Metrics vary over time in a consistent way. All items rated highest in the rst

month, with good initial feedback; then ratings all dipped in the second month.

As issues were identied and addressed, ratings increased again in the third

month.

■ Other considerations impacting usability

— Interaction of the exoskeleton with PPE: coveralls, fall harnesses, tool

belts, hard hats

— Extent of integration: single vs. multi-user (shared)

— Maintenance needs: end-user, tool cribs, exoskeleton vendors

21Proceedings of the 2018 Ergo-X Symposium

— One speaker highlighted the importance of social factors. e unwanted

attention of using systems aected usability most. Even if objective metrics

(time to task completion) improve, employees will not wear the systems if

they receive unwanted attention or comments.

— Quality of life, employee engagement

■ Application of human factors engineering (HFE) of medical devices is helping

medical device manufacturers address the regulatory imperative (that devices are

safe and eective) and the commercial imperative (that devices are ecient and

satisfying).

■ HFE is important because people make mistakes. FDA wants manufacturers

to gure out the ways in which people can make mistakes with a device. is is

known as risk or hazard analysis.

■ Manufacturers should mitigate risks. is should be design-based mitigation or

instruction/information based (such as labeling). Residual risk should be brought

to a minimum, through testing; no product can be completely risk free.

■ Formative testing: identify all potential use-related hazards (risk analysis).

— Identify all critical/high-risk tasks performed by users.

— Identify the interface’s strengths and weaknesses.

— Identify potential use errors that could result in serious harm.

■ Summative testing: perform a human factors validation study to demonstrate

that the device design supports safe, eective use.

— To be representative of the intended users, include in the sample at least

15 users per distinct use group (that is, people who will use the device in

dierent ways, such as patients, caregivers, and therapists).

■ e FDA expects testing with a representative group of users (per age, gender,

occupation, and experience using the device). e user group may be distinct if

they use the device in a dierent way (for example, the physical therapist will use

the device dierently than a patient).

■ Validation testing should contain at least 15 participants per distinct user group.

■ User-group characteristics to consider include these:

— Patient—level of paralysis, enrollment in a rehabilitation clinic, level of

independence/role of caregiver in their life, anthropometric prole

— Caregiver—roles/responsibilities in patient’s life; recruit actual caregiver

or representative substitute

— Physical therapist—methods and any other assistive equipment they use

with their patients

■ Representative use scenarios: e FDA expects that the tasks participants per-

form are representative critical, high-risk, dicult, frequent tasks. Tasks should

follow natural workow.

Proceedings of the 2018 Ergo-X Symposium

■ Consider the percentage of patients who can complete tasks on their own, their

reliance on caregiver, and other physical limitations and safety concerns prevent-

ing them from completing the task.

■ Representative environment: e FDA expects consideration of elements that

might aect use of the system, such as facility layout, accessories, furniture,

lighting level, sounds or distractions, who is present, and access to help. For

instance, is the system typically donned/doed in the home environment or

physical therapy clinic? Who is typically there to assist the user? e researcher

must include realistic performance-shaping factors.

■ Representative device—e FDA expects researchers to ensure that testing

is comprehensive and includes pre-use and post-use (product maintenance).

Validation testing requires that the device is comprehensive relative to the nal

product; all parts are product equivalent; and the eects of using prototypes in

early stage research are minimal. Usability testing is not ecacy testing.

■ In formative evaluation, aesthetic and comfort preferences are commercial

imperatives.

■ Outcome: No errors or problems with use should occur that could cause serious

harm.

Q&A/Discussion

■ Some users initially gravitate to exoskeletons because of a “coolness” or “new-

ness” factor. is initial enthusiasm can drop o in early use before increasing

again. In the military, the culture is dierent. When it denes a system as “mis-

sion critical,” personnel tend to adopt the technology even if they dislike it.

■ Eects on balance were dicult to assess in laboratory representations of tasks,

particularly those representative of ascending/descending work platforms.

■ A conference participant asked about perspectives on making devices mandatory

for certain jobs, on the basis of determined criteria. One opinion was that the

systems have not matured enough nor existed long enough to be treated similar-

ly to PPE.

■ Subjective approaches are prevalent in usability testing. A participant asked

about objective measures for assessing usability. e speaker mentioned that

metabolic cost has a direct relationship with usability and that a human factors

evaluation includes assessing risk and identifying high-severity risks with poten-

tial for harm. Evaluations should establish tasks in which risks can be confronted

for the purpose of identifying user mistakes. Error rates are a measurable way to

determine residual risk.

■ A participant asked about assessment of fall risks. Focus groups with physical

therapists have raised this issue. For industrial use, laboratory tests of fall risks

are challenging because of the need for overhead harness systems.

23Proceedings of the 2018 Ergo-X Symposium

Session Title

Research Methods 3—Assessing Safety Panel

Moderator

Brian Lowe, National Institute for Occupational Safety and Health

Speakers

Ian Marcus, U.S. Food and Drug Administration (FDA)

Assuring the Safety of Medical Exoskeletons: An FDA Reviewer Perspective

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Safety_Ian_Marcus.pdf

Roger Bostelman, National Institute of Standards and Technology (NIST)

Toward Standard Test Methods for Exoskeletons

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Safety_Roger_Bostelman.pdf

Angela Boynton, U.S. Army Research Laboratory

Assessing Safety of Physical Augmentation Technologies for the

Dismounted Soldier

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Safety_Angela_Boynton.pdf

Summary Points

■ e FDA discussed how the U.S. medical exoskeleton regulatory framework is

focused on patient access to high-quality, safe, eective medical devices of public

health importance.

■ e FDA reviewed what the Agency considers a medical device and when FDA

requirements may apply. A medical device is dened by technology and intended

use. e current federal regulation for powered lower-extremity exoskeletons is

21 CFR 890.3480.

■ e FDA discussed the wide range of potential risks that aect the overall safety

of medical exoskeletons.

■ e FDA advocated that medical exoskeleton developers should feel free to

contact them with questions and/or concerns and to submit a pre-submission for

feedback prior to initiating timely and costly testing.

Proceedings of the 2018 Ergo-X Symposium

■ NIST discussed the development of the Position and Load Test Apparatus for

Exoskeletons (PoLoTAE). Others can easily duplicate this wall-based set of tests

to do their own testing.

■ NIST discussed the development of artifacts for use in measuring the movement

of an exoskeleton with an optical tracking system.

■ NIST is developing six task-based tests and one knee test. Testing with volun-

teers (5th to 95th percentile) from NIST is ongoing.

■ e U.S. Army Research Laboratory (ARL) has been working in the exoskeleton

space as an organization for over 20 years.

■ ARL was very involved in the DARPA Warrior Web program and evaluated all

the exoskeleton prototypes.

■ ARL is looking at dismounted-soldier applications for exoskeletons and is using

a safety-assessment approach in its testing methodology.

■ Soldier applications dier from industrial or medical applications of exoskel-

etons. e safety assessment must take into account dynamic environments,

dierent ranges of motion, high-paced activities, and challenging temperatures

and terrains.

■ Stability is a key measure for exoskeleton testing.

Q&A/Discussion

■ Panelists discussed stability as a measure and combining dierent metrics to

address specic stabilities, such as medio-lateral stability.

■ Panelists discussed future test methods for hip, shoulder, and elbow. Making the

data public could enable others to try dierent analytical techniques.

■ Panelists discussed the goal of taking the mentioned test methods and introduc-

ing them into the standards development process through the ASTM Committee

F48 on Exoskeletons and Exosuits.

■ Panelists discussed data sharing across the military, industrial, and medical

exoskeleton application areas. Data sharing would be facilitated by developing

generic test method standards that cover common tasks and requirements.

25Proceedings of the 2018 Ergo-X Symposium

Session Title

Research Methods 4—Assessing Ergonomics

Moderator

Cathy White, Dow Chemical Company

Speakers

Marty Smets, Ford Motor Company

Perspectives on Implementing an Exoskeleton Program in Automotive

Manufacturing

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Ergonomics_Marty_Smets.pdf

Maury Nussbaum, Virginia Tech

Lab-Based Assessments of Occupational Exoskeletons: Overview of

Methods and Results

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Ergonomics_Maury_Nussbaum.

pdf

Summary Points

■ Assessing an exoskeleton’s eectiveness, in terms of reducing physical demands

in a laboratory setting, involves challenges:

— Task context (simulated industrial work vs. basic motor tasks)

— Relevant independent variables that can interfere with or inuence

user-exoskeleton interaction

— Relevant and feasible dependent measures to monitor

■ Certain task-related factors can be manipulated in laboratory simulations:

external loads, tools, postures, workstation congurations, precision demands,

and work patterns (for example, pacing or duty cycles).

■ How do we account for familiarization and learning eects, and hence, what

are acceptable test durations? How do we account for adjustability in some

exoskeletons?

■ Does every device and every new iteration of a device need to be empirically

tested? Are there better approaches?

Proceedings of the 2018 Ergo-X Symposium

■ Dependent variables currently under study are muscle activations, subjective

perceptions (such as discomfort), joint kinematics, kinetics, task performance,

usability, joint range of motion, metabolic demands, postural control and slip-

trip-fall risks, interface pressures, coordination, dynamic stability, endurance

time, fatigue, and model-derived estimates of strength and spine forces.

■ Which of these measures are most useful (and feasible) for guiding decision

making? How do we resolve inconsistencies between measures and/or studies?

■ Fatigue and performance may be the most important to consider, as we move

forward.

■ We need to understand both anticipated and unanticipated outcomes of

exoskeleton use.

■ What is the role of lab-based evaluations in this eld, given that industries have

embarked on their own eld trials based on their specic use cases?

■ An approach for eld implementation of exoskeletons is to treat use of passive

exoskeletons as solely for increasing endurance rather than augmenting strength

and not to use them for speeding up the return to work or assisting a restricted

worker.

■ Fit and functionality testing by actual industrial operators at Ford helped in rede-

signing early arm-support exoskeleton prototypes, in terms of increasing ROM

and device adjustability, regulating thermal comfort, etc.

■ A follow-up phase of exoskeleton eld-testing among workers performing over-

head assembly work showed high acceptance by operators and lower discomfort.

■ Ongoing work will assess eects of long-term exoskeleton use on reduction in

operator discomfort and shoulder injury risk.

Q&A/Discussion

■ Fatigue and performance were highlighted as important measures (for industry

decision makers), but metabolic cost was not highlighted. It was suggested that

metabolic demand may be an indirect indicator of fatigue but may not be related

to injury risk or performance.

■ Ergonomic assessment is dicult in the workplace. We need to perform pre-

dictive analysis. A participant asked whether exoskeleton technology should

be treated as PPE and rolled out as such, and when to go from voluntary use

to mandatory use. We need more research on these topics, not just to mitigate

risk factors but also to understand long-term and cumulative eects. We need

better data and standardization. We also must navigate the political landscape

in unionized workplaces with a focus on lowering injury claims and on lost-time

compensation.

■ Because of complications with hygiene (due to sharing) and time spent in tting,

Ford’s approach is to use exoskeletons individually in the current trial.

27Proceedings of the 2018 Ergo-X Symposium

■ When asked about psychological implications, the speaker responded that the

study data are too preliminary to determine whether users become emotionally

attached to exoskeletons and to comment on implications.

■ e Ford-VT study involves only North America and uses standardized

questionnaires. Some things are managed globally and others by the local plant.

A trained team instructs every site coming into the trial now. e focus is on

establishing and raising level of awareness of these tools and paving the way for

large-scale implementations as sensors and devices get better.

■ Many exoskeleton prototypes are being researched, but only shoulder

exoskeletons are currently under study for production implementation.

Proceedings of the 2018 Ergo-X Symposium

Session Title

Closing Discussion Panel

Moderator

Cindy Whitehead, U.S. Navy—Naval Sea System Command

Speakers

Ben Petro, U.S. Ofﬁce of the Secretary of Defense (Presented by Cindy

Whitehead)

DoD Views on Exoskeleton Development and Use

Delia Treaster, Ohio Bureau of Workers Compensation

Some Thoughts on Industrial Exoskeletons from a Worker Compensation

Perspective

Gerard Francisco, TIRR Memorial Hermann Hospital

Rehabilitation Application of Wearable Exoskeletons

View slides for B. Petro, D. Treaster, G. Francisco: https://higherlogicdownload.

s3.amazonaws.com/HFES/42fbb4-31e1-4e52-bda6-1393762cbfcd/UploadedIm-

ages/Closing_Whitehead_Treaster_Fancisco_closing.pdf

Donald Peterson, Northern Illinois University/ASTM Committee F48 on

Exoskeletons and Exosuits

Exosystems Testing, Validation, and Standardization

View the slides: https://higherlogicdownload.s3.amazonaws.com/HFES/42fbb4-

31e1-4e52-bda6-1393762cbfcd/UploadedImages/Closing_Don_Peterson.pdf

Summary Points

■ Cindy Whitehead presented prepared comments from Dr. Petro, U.S. Oce of

the Secretary of Defense.

■ e insurance industry is numbers driven. e number and the cost of claims are

important.

■ e State of Ohio is monopolistic for workers’ compensation. is means that

employers are insured by the state or are self-insured; there are no private insur-

ers for Ohio workers’ compensation.

■ e manufacturing environment and work tend to be more predictable than

other industries such as public employment and construction employment.

29Proceedings of the 2018 Ergo-X Symposium

■ Postural demands are prevalent in construction work. Posture can be con-

strained by the nished environment and restrictions in how materials can be

carried/handled in nished environments. Technology assistance is necessary in

these applications.

■ e back and shoulder represent the #1 and #2 body part injuries in terms of

costs to workers’ compensation systems.

■ Manual patient handling in the health care industry is a problematic task and

signicant source of injuries. Handling of nursing home residents is a special

OSHA emphasis area because of high injury rates. Current patient lifting/han-

dling aids have shortcomings. ere is a need for exoskeletons designed speci-

cally to assist in patient handling/transfer.

■ Wearable robots improve the services that rehabilitation professionals provide to

patients and assist in delivering the appropriate therapeutic dose. ese profes-

sionals’ perspective on exoskeletons is that they provide assistance to therapists

but are not a substitute for them.

■ More frequent and longer therapy sessions may be more feasible and cost

eective to deliver with augmentation from wearable robots.

■ One hundred thirty-ve countries are represented on ASTM standards com-

mittees. Some countries put consensus standards, such as those developed by

ASTM, into law.

■ Companies are looking at automation of people and their interactions with

machines.

■ Strength-augmentation exoskeletons have been constructed from o-the-shelf

parts for just a few hundred dollars. e low cost is advantageous for

disposability.

■ Performance thresholds for these low-cost devices are still unknown. is

suggests the need for standards.

■ Modular systems may have various congurations, depending on anthropom-

etries and accessories. We need to know how to match the standard(s) with

modular systems.

■ ASTM membership has benets to both faculty and students.

Q&A/Discussion

■ Panelists described the benets of participation in relevant organizations bring-

ing together end users, developers, and academics, such as the ASTM Commit-

tee F48 on Exoskeletons and Exosuits and the WeaRAcon association.

■ A less expensively produced exoskeleton is in development for use in pediatric

and adult populations.

Proceedings of the 2018 Ergo-X Symposium

■ e Ohio Bureau of Workers’ Compensation is sponsoring basic

exoskeleton research in industry environments, by universities in Ohio.

■ e Department of Defense will focus on close-combat military

applications with signicant ballistic protection.

■ WearRAcon, Wearable Robotics Association (Europe), and ASTM F48

committees are fertile grounds for collaboration on this topic.

is page intentionally blank.

Promoting productive workplaces through

safety and health research

DHHS (NIOSH) Publication No. 2020-102

DOI: https://doi.org/10.26616/NIOSHPUB2020102