Human process is painstakingly slow
- By Steven Pisciotta
- Nov 01, 2015
Employee mustering at industrial sites
or large facilities has typically been difficult
to efficiently manage due to the
complexity of the situation and human
factors. Traditional muster plans specify
primary and secondary muster locations. During a
muster drill or actual emergency, workers proceed to
the well-defined muster locations. Team leaders carry
a personnel list and check names off the list for those
individuals during roll call at each muster site. Once
the team leaders merge their lists, they arrive at the final
list of individuals whom were unaccounted for, if any.
Given the human element of traditional mustering
methods, this process is painstakingly slow and
the final result is error prone. The consequences of
this approach can be catastrophic on two levels. First,
critical time is wasted determining which individuals,
if any, are still inside the facility. If individuals are still
inside the facility, critical time may have been wasted
prior to the search and rescue efforts. Second, if the
list incorrectly indicated that individual(s) are still inside
the facility, this will place First Responders into
unnecessarily dangerous situations. Worse yet, a list
incorrectly may indicate that a person has left the facility,
while in fact they still need to be rescued.
Know Who is in the House
Also important are site visitors, contractors or employees
from other facilities. Does the roll call contain
a list of these infrequent individuals? Do these
individuals realistically know the emergency procedures,
and more specifically, the locations of the
Further complicating the situation is the exceedingly
diverse and multicultural nature of the workforce,
which can be quite transient. These factors increase
the potential for misunderstandings during the
roll call. All these aforementioned issues complicate
the situation, not to mention elevated stress levels and
time restrictions in an emergency only exacerbate the
problem and the risk.
Assuming the muster process is performed flawlessly,
the roll call only provides a list of missing individuals.
The search and rescue team needs vital location
information to be successful. What is the last
known location for each individual? What building?
What floor? What area within the floor? This information
is essential for a successful search and rescue
operation but is even more important for a facility
with multiple multi-level buildings.
Regardless of their complexity, all mustering systems
generate three lists: the roll call or inventory
list, the mustered list and the missing personnel list.
How these lists are generated as well as the frequency
of update varies with the application. On an oil platform,
the roll call list is relatively constant during the
workday as compared to the list at an industrial site
where thousands of employees are potentially able
to leave the facility at any point in time. As such, the
site environment greatly affects the real-time requirements
for generating the roll call list, and has a tremendous
impact on the muster system requirements.
For example, at large industrial sites, individuals can
bypass specified muster locations, are completely out
of danger, yet still be considered in danger, simply because
they do not proceed to the muster location.
Modern mustering systems, known as Personnel
Accountability and Mustering Systems, are a system of
systems, which are comprised of three components for
generating and maintaining the aforementioned lists.
The Personnel Inventory System generates the roll call
list based upon each individual carrying or wearing
some sort of tag or badge inside the facility. The granularity
of the roll call data can vary greatly, based upon
the number of tag or badge reader locations. For instance,
it could be as simple as acknowledging when a
person is inside the building or as complex as knowing
a specific zone within a multi-level building. The muster
system generates the mustered list based upon the
individuals exiting the buildings and entering muster
locations or zones. The remote server application generates
the missing personnel list from real time analysis
of the roll call and mustered lists.
In contrast to the traditional model, the design
of modern mustering systems requires a comprehensive
understanding of the worst case catastrophic
scenario. While performing this worst case scenario
analysis, certain notions must be considered about
the personnel and the facility, as well as catastrophic
One aspect to consider is that if people are required
to physically “check in” or register, how likely are
they to actually perform this task? During the chaos
of an emergency, people tend forget procedures or
are inadvertently missed during the muster roll call.
This results in a potential source of list inaccuracies,
regardless of whether it is the roll call list or
the mustered list.
It is imperative to consider the impact of the catastrophe
on worker actions. During an emergency,
workers will resort to just about anything to escape a
catastrophic situation and may not proceed to specific
muster stations. For instance, workers may instinctively
go to a safe location and immediately contact
loved ones or they might just go to their homes. In
either case, they might bypass the specific muster stations
and thus would be included on the missing personnel
Additionally, the catastrophe’s impact on a given
facility creates two crucial assumptions. First, one
has to assume that existing doorways and normal
passageways are inaccessible and workers may exit
the facility via non-conventional methods, such as
through windows. Second, there may be a complete
loss of power and/or damaged infrastructure at the
site. Since the Personnel Inventory System must be
located within the facility, one has to assume that updates
to the roll call list may be incomplete or missing
once the emergency commences.
Design Based Upon Assumptions
These complex factors and assumptions creates real challenges for system designers.
The primary issue is ensuring that the Muster System and the Remote Server
Application are fully operational after the emergency and that the Muster System
provides 100 percent coverage around facility. Minimizing or eliminating the need
for workers to “check in” or register with the Personnel Inventory System and the
Muster System is an important consideration.
Assuming power is lost and the local facilities are damaged requires several
unique design decisions. The first decision is that the Remote Server Application
must be hosted off-site, either by a 3rd party or at another office which does not
contain dangerous or combustible materials. Given today’s technical knowhow
this is a relatively simple task, since most companies have in house IT departments.
As an added benefit, the content can be browser based which would allow the data
to be viewed and disseminated to multiple people simultaneously, regardless of
their hardware platform, such as a laptop and/or smartphone.
The second decision is that the muster system must be able to operate off the
local grid, in terms of both power and connectivity. This could be as simple as
a primary power and/or solar power plus battery backup. Since the local infrastructure
might be lost, the Muster System must also be off the local communication
grid. One potential solution is to use wireless modems, either GSM/GPRS,
CDMA or Satcom devices, which requires a monthly service charge per device.
Depending on the size and nature of the event, these public wireless networks may
be overwhelmed as people try to contact family members and friends. Another is
to use a long-range point to point radio system, which avoids a monthly charge.
Either method will allow the muster system to communicate to the remote server
application without local infrastructure.
Since the remote server application must be located off site, the next decision is
ensuring connectivity to the Personnel Inventory System. The best option is TCP/
IP, which permits continuous real time updates. Modern facilities are typically
wired with Ethernet connectivity or can be retrofitted with Wi-Fi. This would also
allow easy scalability if increased granularity is deemed necessary in the future.
The Personnel Inventory System would be placed behind the local firewall and the
communications to the Remote Sever Application can even be secured via Virtual
Private Network (VPN).
Getting Enough Coverage
The final issue to discuss is ensuring that the Muster System has 100 percent coverage
around the facility. Installing hundreds of short-range muster stations, where
workers are required to ‘check in’ is not feasible, due to the high cost. Installing
several long range muster stations, using either Bluetooth Low Energy (BLE) or
active RFID tags, would be the best solution. In either case, the muster readers
could be placed 50-200 meters apart, depending on the technology and environment.
As the workers pass through the muster regions, the stations read the tags
and relay that information in real time to the Remote Server Application. The
benefit of using an active RFID tag based system, is that it communicates with
passive RFID readers, which can be installed with overlapping coverage for system
redundancy, greatly increasing the reliability of the muster list.
These same readers can be adjusted for short range use in the Personnel Inventory
System. The end result is that the same technology can be configured for
long range monitoring to optimize the muster system with redundancy and short
range monitoring to enhance the positioning accuracy of the Personnel Inventory
System. Thus, by eliminating the check in burden, the entire process can be more
effective and completely transparent to the worker.
By leveraging current technology and considering complex factors and possible
scenarios, system designers can create robust mustering solutions. These systems
decrease the likelihood of error by reducing manual steps, while
being scalable and flexible for future growth. In the end, consideration
of these factors enhances the final product and provides
tremendous safety value for all stakeholders.
This article originally appeared in the November 2015 issue of Security Today.