BBS & HOP = Predictive-Based Safety

With accuracy rates as high as 86 percent, predictive analytics have helped organizations save lives.

• By Chuck Pettinger
• Dec 01, 2018

The Safety field has made continual progress over many decades; however, many organizations have reached a plateau and are looking to move from great to world-class. Companies have robust training and effective safety processes, yet struggle making continuous safety improvements. Great organizations also implement processes like Behavior-Based Safety (BBS) and Human Organizational Performance (HOP) to supplement their compliance-driven programs, yet people are still getting hurt. Although, some people claim that "HOP is the archenemy of BBS" and have set up BBS as a "strawman" to HOP. This article will discuss the next safety evolution will occur through a synthesis, not division, of BBS and HOP using system analysis and predictive analytics to create a more proactive "Predictive-Based Safety" process.

Behavior-Based Safety (BBS)
To reduce the number of injuries, organizations use audits or inspections to judge compliance with their policies and procedures. BBS focuses more on observable safety-related behaviors, rather than on whether people are following rules. Thus, a robust BBS process does not replace audits or inspections; it simply looks at safety from a coaching versus policing perspective. Once a risky behavior is identified, employees analyze system and environmental factors influencing the risky behavior, brainstorm improvements, and then use future observations to verify those changes. Focus on behavior does not place blame but helps identify where organizations might be drifting into potential failure. Thus, observing someone doing something risky helps the employees examine the presence or absence of safety "barriers," not just on catching someone doing something wrong. Finally, BBS focuses more on the coaching relationship between employees, and stresses the importance of the conversation, not just fault-finding.

The origins of BBS stem from Applied Behavior Analysis (not from Heinrich, as often misquoted). This discipline focuses on organizational aspects directing and motivate peoples' safety-related behaviors. BBS began with the Komaki's (1978) study of safety-related performance feedback, improved upon with research by Geller, who coined the term "Behavior-Based Safety" back in the 1980s.

Employee engagement is at the heart of successful BBS processes. However, even with the best implementation, many processes struggle to maintain their momentum. This can be caused by lack of leadership engagement, lack of resources for the BBS team, not using their observation intelligence, or more typically, no sustainability plan. With organizations asking for increases in efficiency, without the BBS practitioner providing a value proposition to their leadership or fellow employees, these processes begin to lose their internal support. To improve your BBS process, practitioners need not start from the beginning, they simply need to evolve their process.

Human Organizational Performance (HOP)
In 1982, on the heels of BBS, clinical psychologists in the UK began their research on cognitive failures and classifying slips and errors. This research was then followed by Human Error founders Rasmussen and Reason. Rasmussen (1986) further classified types of human error into knowledge, rule, or skill-based. Reason (1990) gave his "swiss cheese model" to illustrate layers of defenses with the holes representing latent weaknesses in our processes and systems. Reason hypothesized that incidents occurred when all the holes (latent weaknesses) are aligned. Thus, the field of Human Error focused on identifying those weaknesses and closing the identified gaps. Dekker in 2002 gave us his "new view" where we should not ask who failed, but what failed. Dekker focused on understanding the situation, context, systems and processes that facilitated the errors, and how organizations can "drift into failure" (2011). Over time, the name "Human Error" was changed to a less fault-find and more positive broader label of Human Organizational Performance (HOP).

HOP tools include methods to identify latent weaknesses by looking at learning teams, pre/post-job briefs, (pre)accident investigations, and error-likely situations. One of major contributions is teaching organizations how to react to failure and get out of the blame, shame and train mentality. Much like BBS, if sustainability strategies are not used, HOP initiatives can lose momentum and become the "flavor of the week." Many HOP initiatives fall into the "train and pray" methodology without institutionalizing a process to keep the momentum going. In this case, HOP initiatives track nothing to demonstrate their value—although some HOP initiatives do collect a lot of data but again struggle putting that valuable data to use. There seems to be an opportunity for a synergy, not division, between these two well-established effective strategies that address the shortcomings of both processes.

A Case for Predictive-Based Safety
With the obsessive use of smartphones, tablets, high-power networks, and social networking, safety professionals have more access to ground breaking Internet-of-things (IoT) technology than ever before. However, few safety departments take full advantage of this technological edge. This advantage is especially relevant to BBS processes that produce a large quantity of data. If you look at any recent newspapers, magazines, or websites, you will most likely find a reference to "Big Data," "Predictive Analytics," or "Business Intelligence." Predictive analytics is the study and use of large data sets to predict or forecast. Business intelligence and predictive analytics will be a part of everyone's jobs, if it is not already. And, there is no other area in businesses that could benefit more from using "Safety Analytics" than our safety departments.

It is possible to use your data to predict where your next injury will occur, if you use your data. As organizations upgrade their processes to Predictive-Based Safety through improved technique and real-time data analysis, they will be better equipped to act on their safety analytics. Ideally, using leading and lagging indicators, paired with the power of modern analytics, will help us understand patterns to identify where the next incident, near miss, or critical failure could occur. With accuracy rates as high as 86 percent, predictive analytics have helped organizations save lives.

The safety field collects a plethora of data, from safety observations to near misses. Unfortunately, this critical safety intelligence is often not used, is misused, or is just plain ignored. So our challenge is to use safety analytics to help us predict latent weaknesses in our processes and systems and proactively use the data to create resilient organizations, help them fail safely, and predict and eliminate incidents. To make this happen, our organizations need to create a process for acting upon the intelligence gained from BBS and HOP tools. Thus, the need for Predictive-Based Safety:

1)    Leadership Engagement: During a scheduled weekly leadership meeting, safety analytics should be shared—not by safety, but by operations. The safety professional can help provide the actional information, but the leadership should own the data.

2)    Week-look-Back: Much like an emergency room physician, our leaders need to use our observational intelligence to check "vitals" and "diagnose" potential concerns. BBS provides insights into how activities are being performed, while HOP tools help clarify the context of the actions observed. During the meeting, operations shares last week’s data such as open issues that have been successfully addressed, incidents/injuries, and any other leading indicators, such as stop-work or good-catches. This demonstrates the value of the data being collected and shares our safety analytics.

3)    Week-look-Ahead: Next, the team focuses on closing open issues, error-likely situations, serious incident precursors, critical risk management assessments, or other risky behaviors identified for the upcoming week.

4)    Focus Your Observations: Once the week-look-back/ahead is shared, the team "prescribes" their focus for the week. If there are potential error-likely situations or other critical risks being performed, then we focus on those for the week. If there have been concerning trends, we would focus on that target. These discussions help leaders learn how to react appropriately to system/process failure and not be too quick to place blame on the employee.

This crucial step provides two important roles to process sustainability: It brings value to the safety analytics being collected and provides a different weekly focus to avoid "observer complacency." This drives home not only the "why" we are doing these safety activities, but focuses on "how" we are using BBS and HOP to help keep our organizations safe.

5)    Celebrate the Data: Once the weekly focus is identified, the safety analytics are shared. Teams also should provide positive safety changes. This should not be data overload, but instead targeted information relevant to employees. Fixes also should be shared to drive home the value of the processes. Again, it is all about safety conversations being openly shared and discussed.

Over many years, organizations have seen a variety of strategies and tools used to help keep people from getting hurt. To avoid the "flavor-of-the-week" mentality, companies should not simply move to the next "safety fad," but should instead focus on how to use those philosophies together to capture as much information about processes and systems as possible. This synergy and focus on proactive safety analytics can help organizations realize that safety is not just the absence of injuries, but the presence of defenses. As organizations look to the next four decades, we will no doubt see better use of safety analytics through predictive-based safety and fewer people going home hurt.

This article originally appeared in the December 2018 issue of Occupational Health & Safety.