Safety Management
Six Sigma
Safety
Applying quality management principles
to foster a zero-injury safety culture
By Michael M. Williamsen
I
IS SAFETY GIVEN THE SAME COMMITMENT as
product quality? Are employees accountable for
their own safety? Is safety excellence embedded into
the company psyche? These fundamental questions
are driving todays safety revolution.
conventional quality concepts is its focus on com-
municating measurable error ratios. By incorporat-
ing customer-focused objectives and metrics to drive
continuous improvementand by establishing
processes which are so robust that defects rarely
In much the same way quality management occurSix Sigma quality objectives aspire to reach a
made significant strides during the 1980s, industrial three-parts-per-million error ratio at a 99.9996 per-
safety is poised for its own transformation. This arti- cent incidence. Statistically, Six Sigma variations are
cle provides an actionable approach to how a zero- the standard deviation around the mean, represent-
injury culture can be driven by adopting the same ed by the Greek letter sigma ().
tools and tactics of product qualitys Six Sigma Todays Six Sigma quality community includes cer-
methodology. It includes a previously unpublished tification that incorporates formal instruction, per-
case study that documents the teamwork, method- formance standards, and applying a wide range of
ology and results of a corporate continuous analytical problem-solving tools such as Pareto charts,
improvement team at Frito-Lay Inc.; it involved 40 process maps and fishbone diagrams. Its mastery bor-
plants and 10,000 employees. rows martial arts vernacular (e.g., black belt, sensei) to
Six Sigma tools are nonproprietary, with a grow- define levels of understanding and performance.
ing number of documented references to their statis-
tical origin (ReVelle). This article documents their Six Sigma Control Levels
practical application to safety and their resulting In the authors opinion, what Six Sigma did for
injury breakthroughs (as illustrated in the case study
quality is about to occur in industrial safety. The
and accompanying figures). same desire to eliminate product mistakes is at work
to reduce injury rates. This parallel journey has six
Safety Performance Culture levels. Each sigma control builds on the previous Michael M. Williamsen,
Like all innovations, Six Sigma encompasses the level until the sixth sigmaa zero-injury culture Ph.D., is a consultant with
perspectives of leading thinkers in manufacturing is attained. CoreMedia Training
and production. Although the concept originated Solutions, a Portland, OR-
with a group of Motorola engineers during the mid- One Sigma Control based safety products and
1980s, Six Sigma encompasses the theory and logic One sigma is set in the era of the three Es of services company.
of quality pioneers such as W.E. Deming, Joseph safety: engineer, educate and enforce. The tools for Williamsen has more than
Juran and Philip Crosby to address the question, Is these rudimentary safety mechanics include work 30 years of business
the effort to achieve quality dependent on detecting orders, safety rules, injury investigations and com- change management
and fixing defects? Or can quality be achieved by pliance programs. While barely touching the sur- experience with
preventing defects through manufacturing controls face of why injuries occur, one sigma tools companies such as Frito-
and product design? establish the foundation for creating a safe work- Lay Inc., General Dynamics
At its core, this approach is about improving place. As with one sigma in quality, the perform- and Standard Oil. He
effectiveness and efficiency. Its primary pursuit is ance (conceptually at least) is 68.5-percent earned his Ph.D. in
perfectiona never-ending dissatisfaction with cur- error-free. This level represents the ability to sus- business from Columbia
rent performance. What separates Six Sigma from tain the essentials in worker safety. Southern University.
www.asse.org JUNE 2005 PROFESSIONAL SAFETY 41
The Safety Perception
Survey: 20 Categories
1) Accident investigation. Does your safety system
deal positively with the investigation of accidents? Are
Two Sigma Control the real causes ever covered up for political reasons or to
The tools for two sigma control include observa- meet production quotas? Do employees feel free to dis-
tion programs, job safety analyses and near-hit cuss the underlying causes and circumstances?
reporting. At this level, awareness and analysis tools 2) Quality of supervision. Are supervisors perceived
are applied to reach a two sigma levelor an injury- to be competent in accident prevention? Do they hold
free rate of about 98.5 percent. Research indicates meaningful safety discussions with employees on a reg-
that a 10-percent error level requires about 3,000 ular basis? Do they reward safe behavior?
observations to detect and act on mistakes [e.g., 3) Substance abuse. Are employees with substance
Harry(a),(b); Jackson; Walmsley]. As errors decrease, abuse problems allowed in the workplace? Is there an
more observations are needed to detect the incorrect effective program for prevention and rehabilitation?
activities, which means a one-percent error level 4) Attitudes toward safety. Is there a positive attitude
requires about 10,000 observations to be statistically toward safety at all levels of the organization? Do em-
valid [Petersen(b) 114-118]. It is a benchmark that ployees feel that management is fair and effective in its
underscores how challenging it is to move beyond approach to safety?
two sigma control without adding to the traditional 5) Communication. Do managers and employees
safety repertoire of observation programs and communicate freely on safety issues? Are there informal
rearview mirror reporting. Two sigma safety con- systems of communication in addition to the more tradi-
trol is focused on what is seen in the workplace. tional channels?
6) New employees. Are new employees thoroughly
Three Sigma Control trained in safety? Does training continue on the job with
Three sigma product quality requires well-defined reinforcement from experienced workers?
responsibilities and accountabilities to provide pre- 7) Safety performance goals. Do workers and man-
dictable results on a regular basis. The same is true for agement formulate behavior-oriented safety goals? Are
three sigma safety [Petersen(a)]. Without safety goals effectively communicated to all employees?
accountability at all levels, it is essentially impossible 8) Hazard correction. Is there an effective system for
for a company to attain this level of control. dealing with reported hazards? Is this system understood
Organizations that have been able to move from two and supported at all levels of the organization?
sigma to three sigma generally attribute their success 9) Inspections. Are there regular inspections of all
to the introduction of individual accountabilities into operations? Do employees have an opportunity to par-
their safety programs. Embracing the conventions of ticipate in these inspections?
accountability and personal responsibility is a critical 10) Employee involvement. Are there opportunities
factor in achieving a 99.7-percent injury-free work- for employees to become involved in safety through
place. While three sigma is commendable, companies such means as quality improvement teams, ad hoc com-
at this level still incur lost-time injuries at a rate of mittees or effective supervision?
three per 1,000 employees. Three sigma safety
addresses what is done in the workplace.
across 20 categories (above), cross-tabulated by man-
Four Sigma Control agement, supervisors and frontline employees. The
Beginning in 1979, Dan Petersen teamed with self-administered questionnaire includes 73 questions
Charles Bailey to develop a comprehensive and statis- and provides firms with a statistically reliable method
tically validated safety perception survey on behalf of to answer the questions, Where do our people believe
the U.S. rail industry [Bailey(a),(b); Bailey and Peter- we are weak? and Where do they agree and dis-
sen]. Today, the survey is used to audit an organiza- agree? Todays safety perception survey results can be
tions safety culture and identify perception gaps compared with a database that contains more than two
million respondents. It is a tool
Figure
Figure 1 1 that provides statistically valid
data for industrywide compara-
Action Item Matrix: Accountability Team tive analyses.
This development added an
important dimension to pin-
pointing improvement oppor-
tunities. Not only does it
identify safety shortcomings,
its implementation is recog-
nized as a valuable buy-in
mechanism to set the stage for
continuous improvement work
teamsa necessary component
to reach four sigma control
99.97-percent injury-free. Four-
sigma control concentrates on
the nonobservable what is
believed in workplace safety.
42 PROFESSIONAL SAFETY JUNE 2005 www.asse.org
11) Program awareness. Do awareness programs
stress safety both on and off the job? Do employees look
favorably on these efforts?
12) Performance recognition. Is good safety perform-
ance recognized at all levels of the organization? Are sionals need to implement a similar approach to
workers routinely reinforced on the job for safe behavior what zero-error quality cultures use in manufac-
or is recognition merely relegated to occasional safety turing. To do this, an organizations continuous
awards? improvement teams must own and imple-
13) Discipline. Is the company perceived as taking ment the following:
a fair approach to handling rules infractions? Is the em- A regular, sanctioned
phasis on discipline in proportion to the emphasis on meeting system with action-
positive reinforcement?
14) Safety contacts. Are there regular safety contacts
able rules and mechanisms
and trained leaders to man-
Task: Define
with all employees? Are one-on-one discussions used in
addition to safety meetings?
age the CI process in safety.
Six Sigma analytical
Machine
15) Operating procedures. Are safe procedures seen
as both necessary and adequate by all levels of the
techniques/tools with safety
issues and projectible data.
Operator Role
organization? Are employees actually aware of the Once these critical fac-
companys safety-related procedures? tors are in place, a zero- Definition
16) Supervisor training. Are supervisors perceived to error safety culture can be a The key safety accountabilities
be well-trained and able to handle problems related to recognized strength along- of the operator are to use safe work
safety? Is their performance measured and rewarded side the traditional business practices, use all safety equipment
appropriately? necessities of customer ser- when required and promote safety
17) Support for safety. Is the whole organization seen vice, quality assurance and with coworkers.
as working together to create a safe work environment? manufacturing efficiencies.
Is each level of the organization perceived as contribut- Responsibilities
As the case study will illus- 1) Before each shift, inspect/
ing effectively to the safety effort? trate, the resulting savings
18) Employee training. Do employees feel that they check the work area to identify any
in both cost and hardship unsafe issues and correct or initiate
receive adequate training in how to work safely? Do can be dramatic.
employees understand how to work safely? corrective action as needed.
19) Safety climate. Is the climate conducive to adopt- 2) Perform daily housekeeping
Applying Six Sigma duties to keep/maintain work area
ing safe attitudes and work habits? Is safety perceived as Tools in the Workplace
important to the organization? in a safe and clutter-free condition.
Five and Six Sigma in- 3) Attend and participate in all
20) Management credibility. Is management seen as jury control requires statis-
wanting safe performance? Are they willing to provide shift supervisor safety meetings.
tical process control tools, a 4) Team with the supervisor to
necessary resources to achieve this performance? dedicated continuous im- present/discuss topics in the super-
Source: Bailey(b). provement (CI) team and visor safety meeting (two to four
active participation from all per year).
levels of employees. This 5) Initiate and follow up on safety
latter component empha- work orders.
Five Sigma & Six Sigma Control sizes the importance of effec-
The next challenge is to use the data from the pre- 6) Provide appropriate safety and
tive meetings. Organizing health training to new/transferred
vious four levels of safety: effective subteams to execute
injury and work order data; personnel.
tasks is essential. Furthermore, 7) Review and improve job haz-
observable processes; because many of the subteams
accountabilities; ard analyses regularly.
combine cross-functional em- 8) Be familiar with all documents
information based on a safety perception survey. ployees from disparate groups,
The material from these four areas needs to be in work area.
it is critical to delineate proven 9) Pay attention to coworkers and
applied in a rapid, accurate and functional way. principles to create a meeting
Once a company is nearing four sigma, the major outside personnel working in the
structure that ensures efficien- area. If they are not following proper
barriers to effective cross-functional continuous cy, participation, action and
improvement are eliminated. A roadmap can be practices or procedures, talk with
high performance. them immediately about correcting
developed to an unprecedented five sigma (99.997
percent) and Six Sigma (three injuries per million their activities.
Effective Meetings for 10) Inspect containers to ensure
employees) safety performance. At this point, an Continuous Improvement
organization can approach a zero-injury workplace. that they are labeled correctly. If not,
To achieve results from safe- relabel them immediately.
As in a Six Sigma quality program, all founda- ty meetings, the person who
tional mechanicsengineer, educate, enforce, calls the meeting must focus on Measures of Performance
observe, investigate, accountability principles and its purpose and desired out- 1) Appraisal by supervisor of
thought patternsare required to establish an comes. By deploying the POP individual task achievement.
authentic Six Sigma safety culture. The challenge is modelpurpose, outcomes, 2) Observations by supervisor.
to create a sustainable safety culture where height- processthe group can remain
ened safety decisions occur without thought. It is a focused and on task.
process that begins by addressing the milestones to
continuously improve. Purpose
Good data are necessary. However, to achieve The purpose is a mini-mission statement. Why is
four sigma performance and beyond, SH&E profes- the group meeting? If the purpose is unclear, start
www.asse.org JUNE 2005 PROFESSIONAL SAFETY 43
A Case Study: Six
This previously unpublished case study result in a spaghetti diagram. The team gories include people, methods, machin-
illustrates how Six Sigma measurements then analyzes each step in the process ery and materials. As problem situations
were applied to a Fortune 500 food prod- being studied and optimizes each indi- vary, this Six Sigma tool has the added ben-
uct company that was experiencing hun- vidual task to a point where inefficien- efit of being able to creatively identify dif-
dreds of injuries across multiple facilities. cies, errors, complicated spaghetti and ferent elements to better fit the individual
The initiative resulted in a rapid safety hazards are eliminated. situation. For the food products company,
improvement in workplace injuries and
the start of a zero-injury safety culture. Cause-&-Effect
Diagram
Figure
Figure 3 3
Pareto Charts
The Pareto chart is one of the most help-
As the CI team con-
tinued its efforts to
Injuries by Gender
ful visual tools in the safety Six Sigma tool Are males or females more apt to have costly injuries?
eliminate back and soft-
box. These charts help to pinpoint unac- The pie chart is an effective tool whenever the variables
tissue injuries, the safety
ceptable occurrences that warrant high pri- are limited and the sum is 100 percent. This example
team used another Six
ority. The charts (Figures 2-8) show the reveals a need to find out why so many women were
Sigma tool, the cause-
frequency and severity of problems and getting injured.
and-effect diagram (Fig-
where they occurred geographically.
ure 13, which is also
Process Maps referred to as a fishbone
Process maps or process flow dia- or Ishigawa diagram).
grams graphically illustrate how a task or Team members were
process can be accomplished effectively able to refer to the chart
within the constraints of time and re- to identify multiple
sources (Figures 9-12, pp. 47-48). This tool potential causes for the
allows a continuous improvement team problem at hand. The
to break down a complicated sequence of bones of the normal
events into simple metered steps, which potential cause cate-
Figure
Figure 2 2
Lost-Time Injuries for 10 Periods
A baseline must be determined to indicate where the investigation should
Figure
Figure 4 4
begin. This figure illustrates lost-time injuries over the last 10 months by vari- Lost-Time Injuries
ous departments, and breaks out which departments warrant the most atten- By sorting injury data, the company drilled down on
tion (e.g., packaging with 52 injuries, then shipping and processing). Once back and shoulder/arm injuries for individual facili-
identified, plants with high numbers of injuries in these departments deter- ties. Ultimately, it was found that all the sites were
mined where to begin the continuous improvement in safety initiative. experiencing similar injury patterns, which presented
a high-priority focus area.
Visually inspect
caseno good,
throw out
Unfold case
Apply label
Fold case
Throw case
Figure
Figure 1313
Cause-&-Effect for Packers
Environment Methods People
Air conditioned (cold) Poor training Untrained
Standing No preconditioning Unconditioned
No footrest/cushions No return to work physical No hiring profile
Only assigned breaks Little rotation of workstation Returned injured EEs to same job
Work hurt philosophy exists No assimilation program No work hardening
High turnover of EEs Inconsistent supervision No EE input to improve situation
High turnover of supervisors Training support by management lacking Supervisors at low training level
Supervisers dont reinforce proper procedures Entry-level job Soft-tissue injuries
EEs clean up spillage/area 20-percent long-term EEs $1,500,000/year
EEs responsible for quality control direct cost
Not state of art Doesnt stack for ease of pickup Slippery bags 60 people injured
Auto case packers coming? Isnt ergonomically designed Various sizes
Throw cases Different stacking configurations
Stack cases Different placement configurations
All manual tasks Seals fail regularly
For cases Cases often bad
Speeds seemingly set high
Incorrect heights for EEs
Little to no automation
No diagnostics
Technology Machinery Materials
Conclusion
The case study
and figures demon-
strate how a CI
approach helped to
improve safety per-
formance in a man-
ufacturing setting.
Injury data were
combined with perception survey data to obtain a Petersen, D.(a). Auth-
entic Involvement. Itasca,
full spectrum of workplace realitiesboth observ- IL: NSC Press, 2001.
able and hidden. Hourly and salaried employees Petersen, D.(b). The
then teamusing Six Sigma tools and effective safe-
ty meeting techniquesto develop and implement a
Challenge of Change:
Creating a New Safety
Figure
Figure 1717
zero-injury safety culture, a workplace that neither
tolerates, nor experiences, injuries.
Culture. Portland, OR:
CoreMedia Training Lost-Time Injuries
Solutions, 1993.
Petersen, D.(c). Safety
Management: A Human
Frequency
References
Bailey, C.W.(a). Improve Safety Program Effectiveness with Approach. Goshen, NY:
Perception Surveys. Professional Safety. Oct. 1993: 28-32. Aloray Inc., 1988.
Bailey, C.W.(b). Using Behavioral Techniques to Improve ReVelle, J.B. Six
Safety Program Effectiveness. Washington, DC: American Assn. Sigma Problem-Solving
of Railroads, 1988. Techniques Create Safer,
Bailey, C. and D. Petersen. Using Perception Surveys to Assess Healthier Worksites.
Safety System Effectiveness. Professional Safety. Feb. 1989: 22-26. Professional Safety. Oct.
Harry, M.J.(a) Framework for Business Leadership. Quality 2004: 38-46.
Progress. April 2000. Walmsley, A. Six
Harry, M.J.(b) Six Sigma: A Breakthrough Strategy for Sigma Enigma. Globe and
Profitability. Quality Progress. 31(1998): 60-64. Mail. Oct. 1997.
www.asse.org JUNE 2005 PROFESSIONAL SAFETY 49