- OIL & GAS
The “venomous cycle”
One of the biggest strengths, and weaknesses, of BBS is the enormous amount of data collected. Not using valuable information collected from observations can lead to an early demise of your process.
Although many behaviors are motivated internally, people often look to external measures, such as praise from peers, to gauge their performance as well. Observation intelligence gained from your BBS process can likewise be a powerful gauge as to how individuals are performing, how the process is progressing, and where your next injury might occur.
Too often, behavioral data is not used.
A typical scenario: The newly trained observer dutifully performs their first observation, provides feedback to their coworker and turns in the checklist to be tallied, but nothing happens with their data.
This is the critical point where many processes fall into the “venomous cycle.” If the new BBS observer sees nothing done with information from their observations, they begin to lose faith. Then the quality of their observations suffers and “pencil-whipping” begins. Once the team starts looking at their data, they find it very close to 100-percent safe and start questioning the value and reliability of the BBS observations. The leadership team also looks at the nearly 100-percent safe observations and yet sees employees still getting hurt, and they begin questioning the time and resources put into supporting the BBS process. Once the team starts questioning the value of the data, they use it less often and only use the checklists to count how many observations were done.
The “virtuous cycle”
One method of generating a step-change in your BBS process is by creating a data use plan. As soon as the process starts, put a data use plan into place to ready the organization to act on the observation intelligence. Designate different levels of leadership and other important individuals in the organization. Decide which reports should be distributed to key individuals and at what frequency. Specify what they should do with the data once they get it. The data use plan should identify as many opportunities to communicate the information as possible. Reports should demonstrate the value observations are bringing to the organization (e.g., external consequences).
Sample metrics to demonstrate BBS value could be:
1) Top 5 risky behaviors
2) Number of work orders created
3) Time to closure of open items
4) “Before and after” pictures of changes due to suggestions
5) Percent of risky observations with comments
To reverse this venomous cycle, teams need to make the BBS process valuable to employees and the organization. If employees see their observation intelligence used, the quality of their observations will increase. Once teams and leaders begin to trust the reliability of data, they gain insight into processes and systems putting their employees at risk for injuries. With this increased organizational transparency, leaders can better provide the resources needed to make the workplace safer. In return, employees see the value from their observational efforts and produce even more accurate and honest observations. This virtuous cycle creates interdependency where everyone is working toward a better safety culture.
Increase the color of observations
To create a step-change in your process, BBS teams should gain as much “color” from observations as possible. One method would be to use a simple 2x2 severity matrix (Severity by Probability) to rate risky behaviors (see Fig. 1).
For instance, if an employee is not wearing glasses while doing computer work at their station but still technically inside the PPE boundaries, this may be recorded as a “low” risk behavior.
Whereas a worker using a grinder without wearing eye protection may be scored as “High” or “Life Threat”.
With data having more “color,” the BBS team can start looking for the data to tell a story. For example, if there are only low-risk behaviors, this may point to a culture that is hesitant to record something too risky. If risky observations are only high or life threat, this may point to a “ticket writing” type of observation where people only observe when they see a risky behavior happening and grab a checklist.
Once severity is added to your process, you need to calibrate your observers. Split employees into four groups. Give each group a flip chart and label them low, medium, high and life threat. Have the first group list as many low-risk behaviors as they can. Have the other groups, in turn, list medium, high and life threat behaviors. Once finished, have them rotate to the next flip chart and add, edit or clarify the list of behaviors. Once each group has rotated through each flip chart, you have calibrated employees and have a list that can be shared with all employees to guide their observations. Adding severity to your checklist and developing a list of risky behaviors sends a strong message that recording risky behaviors is not only acceptable but desperately needed.
Safety analytics: The business intelligence of life
Predictive analytics is the study and use of large data sets to predict or forecast. People have experienced the power of predictive analytics if they have ever used a Google search, had an item recommended by Amazon based on previous purchases, or bought insurance. Business intelligence and predictive analytics will soon be a part of everyone’s jobs, if it is not already. And no other area in business could benefit more from using “safety analytics” than our safety departments.
The safety field collects a surplus of data from safety observations to near-misses. Unfortunately, this critical safety intelligence is often not used, misused or just plain ignored. With increasing computing power now accessible to organizations, it is possible to perform complex analytics that could not have been done a few years ago.
The 4 safety truths
Schultz (2012) and his colleagues took a large data set of safety observations and partnered with data scientists from Carnegie Mellon’s Language Technology Institute. The team used analytics similar to those used when they helped IBM create the computer program “Deep Blue,” which beat the grandmaster Kasparov in a series of chess games. With over 113 million safety observations from 15,000 different locations, Schultz and his team discovered four “Safety Truths” through the use of predictive analytics, which were published in a recent whitepaper.
Embrace data & technology to leave a legacy
As the first generation raised on the Internet, Millennials (born between 1980 and 2000) are quicker to adopt new technology and are heavily influenced by social media sites. Also, a larger proportion of this generation uses smartphones (60 percent) than other generations (Barkley 2011). As such, we have a great opportunity to take our BBS processes to the next level by engaging this 25 percent of our U.S. population.
Once we start putting data to use, the more data we will have to use. When we start using safety analytics to help pinpoint where our safety resources need to go, we will need a way to harness this momentum to keep this sense of urgency fresh. To create that step-change, we need to adopt more “social” ways of getting safety critical information into our hands. Using mobile devices to enter observational data is one way to capture Millennials. Using smartphones and tablets also will increase the timeliness of information. To leave a safer legacy for Millennials, we need to embrace technology, act on the data and begin to use safety analytics to predict where our next injury will be.