Every day drivers operate their vehicles amidst the chaos of growing congestion caused by stressed-out and distracted drivers. Vehicles drift out of lanes as drivers distract themselves with gadgets or texting. Other drivers mindlessly tailgate with the mistaken belief that by driving close to the vehicle in front they’ll somehow arrive sooner.
When two (or more) vehicles almost collide, but don’t, many motorists might utter some exclamation, shake their head about how bad the other driver is and continue on their way. Many safety professionals would label the incident as a “near miss”1,2 – as in “it was nearly a collision but we missed”.
Collecting data on near miss incidents in the workplace is an emerging part of many safety professional’s jobs: they have the advantage of making direct observations of the workplace, soliciting feedback from employees and even building a culture of self reporting of near misses.
The driver safety specialist has a harder time obtaining near miss data since drivers are not likely to self report near misses as they happen, make detailed observations of all circumstances, or remember details from an early morning incident at the end of their day.
If there were convenient mechanisms enabling us to collect and analyze near miss data, it would help us:
- Address flawed processes and procedures (i.e. scheduling, routing, dispatch, etc.)
- Investigate enhancements to equipment (i.e. mirrors, steps, rails, etc.)
- Adjust educational programs for content, length, periodicity, etc.
- Cross reference this new data against historical information (i.e. crashes, violations)
- Tailor our finite attention and resources where it can have strongest and most immediate impact.
Most importantly, it would give us an opportunity to hit the “pause button” on life and compassionately intervene with our operators – perhaps before a “real” collision actually happens, injuries are incurred or violations are issued.
Don’t we already accomplish this with historical data?
Fleet safety professionals have tirelessly identified drivers who may be “at-risk” of becoming involved in a collision by looking at historical data. The phrase “historical data” is often interchangeably called “lagging data” since the information lags behind the actual incidents that we aim to avoid.
The difficulty with depending on lagging indicators of performance is that the driver must “suffer” the consequences in order to appear on the safety professional’s reporting matrix.
How would a driver “suffer” negative consequences? Getting cited by a police officer would typically result in:
- a fine paid out of pocket by the driver,
- lost employment time if the driver must appear in court,
- increased personal car insurance for the driver, his/her spouse and any other drivers (i.e. teens) in their household
- decreased employability of the affected driver due to a littered MVR/Abstract which is reviewed by prospective employers as part of their hiring/screening process
- possible loss of driving privileges, suspension or revocation of license
- possible loss of employment
In short, lagging indicators provide very valuable insights, but come at a very real cost to the organization and the employee. Leading indicators of performance, while harder to pinpoint, chart the way forward towards prevention and avoidance.
The comparison of leading vs. lagging indicators led one commentator to ask the provocative question; “Are your managers operating as company doctors or coroners?”7 Put another way, is the focus of your effort principally to increase wellness, or does it feel like you’re spending most of your time doing “post-mortem” examinations?
To be very clear, the MVR/Abstract review process and post-crash investigations, et.al. are vital safety tools and shouldn’t be abandoned; however, it is clear that it would benefit the driver and the employer to find ways to identify “at-risk” drivers before they receive violations or get into crashes.
The identification and inclusion of “leading indicators” (indicators of “at-risk” performance, habits or behaviors prior violations or crashes) would make a significant difference. Near miss reporting would be one set of leading indicators that could help.
How might we get this data?
There are a number of resources available that provide insight into “near miss” events – those events that would have been collisions – IF – conditions had been slightly different, or reports of habits/behaviors that if left unchecked will likely lead to collisions or violations. I’d like to offer a short list of some examples:
- Commentary Drives and Supervisor “Ride-Alongs”
- Driver Safety Hotlines (aka “How’s My Driving?”)
- Tachographs, Electronic On-Board Recorders (EOBR), “Telematics” devices
During commentary drives, near misses may occur which provide the opportunity for immediate “no-fault” training and coaching. This approach is highly proactive, but requires a tremendous amount of resource time to ride with 100% of drivers. Some fleets use this method with all drivers on a periodic basis (once every X years) and others use it when drivers have been identified as needing help through lagging or other leading indicators. This second approach (selectively using commentary drives) reduces the number of drive events to effectively coach those drivers who may be at greatest risk.
Another efficient resource for near miss and behavioral data is the humble and sometimes misunderstood Driver Safety Hotline. Sometimes called a “How’s My Driving?” program, it is designed to solicit feedback on behavior – praiseworthy or risky – from other motorists. Safety Hotlines have been repeatedly studied by groups large and small to see what effect the program has on crash rates: these studies provide compelling testimony that the data leads behavior and management intervention reduces crashes by 10-30% in most cases. These studies were conducted in a manner that is similar to studies validating most technology platforms, and in one case included three and a half years of data derived from 30,000 power units in varied industries from among 200 fleets. This provides much richer statistical data than many technology studies that were limited to “test pilots” of 25 to 50 vehicles in one or two fleets for time periods of less than a year.
The secret of the success of this program has been:
- Using the reports as a springboard for “no-fault” refresher training instead of blame setting9
- Incorporating the reporting as a positive element of safety cultures and behavior safety programs10,11
- Managers who will discuss the report with the driver to set individual safety goals for modified behavior.12
On-Board recording devices generate specific data sets on vehicle performance and by extrapolation, driver behavior. Data sets typically include vehicle location through the course of the day, speed, harsh acceleration, harsh braking, swaying and sudden shocks or bumps. Most systems report this data to a central reporting hub and management can review the historical data for exceptional events. Some systems provide immediate feedback to the driver as events occur: either through a flashing light or some sort of sound making device.
The successes of most on-board devices has been clearly documented in fuel savings, idle-time reductions, man power resource tailoring from routing efficiency and other “operations” metrics. These benefits are significant, but don’t directly impact the reduction of crashes by themselves. The challenge to most managers is finding the best way to translate volumes of data into enhanced behaviors.
You see, telematics data may be generated in a very different manner than a How’s My Driving report or commentary drive, but the application of that data to affect driver change can be as poorly executed or as brilliantly managed in any of these programs.
Telematics data showing speeding events can be hotly denied by drivers who’ll come up with clever (and often accurate) responses – locations are often “estimated” by satellite triangulation, and sometimes speeds are misread based on locations at crossroads or underpasses, etc. After managing the data from both telematics and safety hotline programs (our clients have begun sending their telematics data to SafetyFirst for enhanced reporting and better training options), we have found that drivers are actually more inclined to deny the telematics alerts than the how’s my driving reports!
Additionally, the amount of data from some telematics programs can become overwhelming. I’ve heard safety managers say that “there’s probably a lot of really good information buried somewhere in the pile of reporting”, but they can’t manage it on a daily basis. If you call it “information overload”, “background noise” or even “dial tone” then you’re likely to move on to other priorities or return to only working with lagging indicators.
Our own firm’s experience is that there are clever and easy ways to avoid information overload and get traction in translating the data into a well crafted coaching session.
By sorting the urgently actionable items from the background noise and then leveraging the coaching processes pioneered and perfecting in the safety hotline program, our hybrid approach managed to reduce excessive speeding by 600% in one year at a major fleet operation. (Click HERE)
Fortunately, in the case of commentary drives and driver safety hotlines, the amount of data is self-prioritizing. The ride-along supervisor can prioritize in real time as the drive continues, and most fleets using hotlines only get reporting on about 2% to 3% of their drivers in any given month (focused on the most egregious behaviors seen on the highway – motorists are not motivated to report trivial issues).
Dealing with Data Organizationally
Dealing with raw data, whether we call it a near miss report, motorist observation report or telematics alert, presents opportunities and concerns:
- We ought to be respectful of drivers and their privacy – no one wants to see their personal data on the company bulletin board as an “example to others” (i.e. share the lesson to be learned, but don’t embarrass the operator). Near miss reporting programs can be most productive when conducted in an atmosphere of trust and proactive goal setting rather than couched in threats and blame.
- We need a system to hold and correlate the data – to provide meaningful management reporting that can distinguish patterns and trends that may signal a larger policy/procedure or system issue13
- We need access to urgently actionable data in a timely fashion – to coach drivers while the event is still fresh in their mind14 while suppressing “background noise” data15
- We need to have a data retention plan in place to either preserve data from spoliation or to properly dispose of old records when the data is no longer relevant to our near miss program16
- We need to develop policies and procedures that create a uniform method to dealing with the data — that it needs to be used to educate and redirect behavior – not as a blame setting tool. Playing the “GOTCHA” game with drivers isn’t likely to improve results or encourage them to embrace the technology that is “getting them in trouble”. Working with an individual driver to set personal and professional goals related to changing habits can be challenging, but also lifesaving (or injury/violation avoiding) at the same time.
Driver Attitudes vs. Manager Attitudes
When capturing near miss data, drivers and managers may argue about data quality or what should be done with the data that is acquired.
- In the past, we’ve met professionals who used to argue about “crank calls” on hotlines as a reason to ignore the data; however, 98% of hotline reports are confirmed accurate by safety managers who actually investigate each report and talk with their drivers. The 2% of reporting that is discarded or deleted comes mainly from transcription errors (i.e. transposed vehicle numbers, etc.)
- Even commentary drives are susceptible to data quality errors: no two driver supervisors will share the same biases or spot all the possible hazards given the same route, same vehicle and same day.
- Telematics data, while “scientifically” obtained can also be erroneous and a source of contention for drivers if they feel threatened by its “alerts”.
My point? Any near miss or leading indicator program could become a source of arguments and negotiations, or become a proactive “game changer” in terms of safety results. It’s really up to the management team to decide if they’re going to help drivers improve through a positive coaching and training program or merely spin the revolving door of driver turnover by using data merely for discipline. Unfortunately, it’s often easier to play the “gotcha” game of confrontation over alerts than to actually make the time to have an eye-to-eye, “no-fault” coaching session about improving habits to be safer while behind the wheel.
The first step is training supervisors on how to use the data to get a positive change.
SafetyFirst, in close cooperation with its own clients, has produced a supervisory training program called “Coaching Drivers – Conversations That Make a Difference”. This program helps managers to make coaching sessions a positive experience by keeping focused on the safety lessons to be learned without getting sidetracked into confrontations over blame and who was right or wrong.
Without a consistent coaching process in place, the most accurate leading indicator, or near miss data, will not be effective in getting drivers to change habits.
This isn’t just a good idea – it’s been studied. Researcher Sunil Lakhiani of the University of Wisconsin-Madison presented findings from a study on near-miss reporting systems to the American Institute of Chemical Engineers at the 7th Global Congress on Process Safety in Chicago in March 2011. Lakhiani reported that a positive management culture towards safety made a significant difference in the employee’s participation in near miss programs.
Recently, one of our larger clients inadvertently proved the coaching point as well. They had installed telematics devices in several thousand vehicles for dispatch, routing, and related reasons. During the first full year using the system, they also accumulated 1700+ excessive speed alerts (above 80 mph for a minute or more), but had no mechanism to push the alerts out to the drivers for coaching. We worked with their telematics provider to have the alerts sent to our safety hotline where we treated the alerts like a motorist’s call-in observation. The net change was dramatic. By sending training materials, requiring the location manager to coach the driver and return the completed report showing goals discussed, training completed and corrective actions taken, they dropped the number of alerts to under 200 in the subsequent year. Coaching makes a difference regardless of the data source if it’s handled in an atmosphere of trust and prevention.
What about driver education as a leading indicator?
Some safety professionals have considered driver education (its frequency and aggregate duration) to be a possible leading indicator of performance. Let’s briefly consider the case for and against this conclusion.
Driver education can be used to introduce new skills or remind drivers of practices and procedures that they should already know due to previous educational experience. In the case of new skills, many safety professionals may argue that habits not a lack of driving skills are the predominate cause of (arguably) 90% of all motor vehicle crashes.17 Therefore, the use of education programs to:
- Remind operators of key safety policies affecting their daily activities, and to
- Increase situational awareness and the rapid recognition of hazards while driving
is an ideal practice to help reduce the likelihood of future collisions. Additionally, the assumption that an increase in education events (frequency/periodicity) or the overall number of hours of training (duration/aggregate) can reduce collision rates seems highly reasonable.
While we’ve characterized commentary drives as a near miss reporting platform, they were initially introduced as a method of driver education. In this regard, they may be part of a leading indicator measurement system, too.
Individual fleets may set driver education as a leading indicator to be verified in hindsight (did crashes go down during the year we increased our education efforts?) Indeed, firms who introduce the varied near miss reporting systems already discussed will likely increase their education efforts as they conduct refresher sessions with drivers who participate in commentary drives, get telematics alerts or safety hotline observations.
Unfortunately, this author has not located many detailed studies published on the links between driver education and crash results other than the ATRI study conducted in 2008 titled; “A Technical Analysis of Driver Training Impacts on Safety”18 In this study, 17,000+ driver records were studied to examine correlations between training and collisions/violations:
“The total “contact hours” or hours of interaction provided by training programs vary greatly from 88 hours to 272 hours. In addition to identifying the total contact hours a student is exposed to in a training program, participating training institutions provided details on the number of training hours that occurred within various training environments, such as the classroom, in-truck, behind-the-wheel and using a simulator. These environments vary between programs, with programs weighting and emphasizing classroom and in-truck training differently. Additional information was collected on the type of instruction that takes place within each training environment.”
The report’s conclusions included:
“…the analysis finds little variation among driver safety performance that can be explained by training program duration within the range of 88 to 272 hours.”
“…the lack of a safety improvement trend line towards the longer duration programs does not provide the researchers with a basis for this conclusion.” [that more training would necessarily result in greater safety results]
To be very clear, this author is not suggesting that driver education is any less valuable or critical to a firm’s safety program. In fact, it is crucial. It may be especially valuable when used to focus resources on those individuals deemed “at-risk” by the near miss system. As a leading indicator, it may be more valuable when blended with near miss reporting or other elements of an existing driver safety program.
Setting a strategy for success
Hopefully you are feeling encouraged that driver safety programs can greatly benefit from incorporating near miss reports into their existing safety program. Near miss reporting serves as a leading indicator to help balance your “scorecard” of valuable lagging indicators such as historical crash reports and MVR/Abstract profiling.
Each organization endeavoring to launch a near miss program should make a plan on how to incorporate this new data into their current safety program:
- Outline where data will come from and how it may be used (i.e. will it be used for education only, or can it be used for discipline, if so, under what circumstances?)
- Develop a process to deal with system faults or physical hazards (i.e. dispatch errors, maintenance items, equipment issues, loading processes, etc.)
- Review historical crash and/or violation data (lagging indicators) and compare to near miss data for trends (i.e. prior to all preventable crashes, these types of near miss reports or leading indicators were present; therefore, if we see these near misses or leading indicators, we need to respond urgently to prevent a crash)
- Have a clear process or procedure – who will be responsible to collect and distribute data?
- Develop a coaching process to interact with affected drivers to affect a change in habits
- Develop a process to track the success of the program (i.e. are collisions decreasing, are police citations/violations decreasing?)
- Celebrate the progress with all affected employees – include them in the results as well as the coaching sessions
Near miss reporting is a valuable tool. It requires a strong, consistent commitment from the management team at all levels to use the data to compassionately intervene with drivers in a trusting manner. Building trust will take time, but it pays huge dividends in safety results for both drivers and management teams. Coaching and education are two sides of the same coin, and each has it’s own supporting role to play. If you want to get on the leading indicator side of the driver safety equation (while not abandoning lagging indicators), then near miss reporting may be the place to start.
During my safety career, I’ve learned that despite all of the networking, conferences and research, I know I don’t have all the answers. I also know that together we can each contribute pieces of the puzzle to get to a better understanding of most any safety issue. I’d love to learn about your experiences with near miss reporting in fleet operations, and hear about your concerns about leading indicators, too. How do you currently identify drivers who may be “at-risk” of becoming involved in a collision? Do you incorporate leading indicators into your driver risk profile, or just lagging indicators? Do you want to launch a leading indicator or “near miss” program, but aren’t sure where to start?
Many safety professionals are active on social networking sites like LinkedIn and share comments and questions through discussion groups. Would you be willing to discuss this article online? If that’s too “public” of an environment, I’d be very happy to talk with you directly, too (1-888-603-6987 toll free).
7 – http://www.bptrends.com/publicationfiles/TWO%2003-09-ART-Leading%20vs%20Lagging-Gotts-final.doc.pdf “Leading Indicators vs. Lagging Indicators” by Ian Gotts, March 2009, BP Trends
10 – http://my.safetyfirst.com/newsfart/ISA%20December%202009.pdf – “Changing Unsafe Behavior Using Activators and Consequences” by Andrew Salvadore, December 2009, Arborist News
11 – http://www.treecareindustry.org/pdfs/EXPO/ABCsOfHumanBehavior.pdf – “ABCs of Human Behavior” by Andrew Salvadore, TCIA Expo presentation
12 – http://vimeopro.com/safetyfirst/safetyfirst-coaching-tutorial/video/30495547 “Coaching Drivers – Conversations that make a difference” by SafetyFirst Systems, December 2011
13 – http://www.ntsb.gov/doclib/reports/2000/HAR0001.pdf NTSB Highway Accident Report conclusion; “…policy of disregarding anonymous calls to 1-800-SAFEBUS prevents the company from identifying patterns of unsafe driving practices by particular drivers or on particular runs and diminishes the potential safety oversight benefits of this program. Including all complaints in driver personnel files would enable [company] to better detect an operator problem and act to eliminate it before an accident occurs.” [italics added for emphasis]
14 – http://www.virtualriskmanager.net/main/aboutus/niosh/t1-2_paul-farrell.ppt “Negligent Entrustment – When is a license check not enough?” by Paul Farrell, International Conference on Road Safety at Work, February 2009
15 – A technology vendor’s presentation states (about their own system) “There are many reasons why a device might trigger: Pot holes; Unpaved roads; Railroad tracks; Turning hard in a large vehicle; Rocking an unloaded tractor-trailer; Waste truck throwing a trash bin into the vehicle hard; Jack-rabbit start; Vehicle Maintenance; Defensive Driving/Evasive Maneuver…” excerpted from http://mcsac.fmcsa.dot.gov/documents/June2010/DriveCam%20presentation.pdf
16 – http://www.atla.org/cps/rde/xchg/justice/hs.xsl/14259.htm “Danger On The Road – The mighty trucking case” by Jeanmarie Whalen, Trial Magazine (American Association for Justice), February 2011, Vol. 47, No. 02
17 – http://www.virtualriskmanager.net/main/aboutus/niosh/t2-3_lynn-berberich.ppt#18 “Crash Analysis and Benchmarking as Tools to Improve Fleet Safety – or – What Metrics Should I Use and How Should I use Them?” by Lynn Berberich, International Conference on Road Safety at Work, February 2009
18 – http://www.atri-online.org/research/results/driver_training_impacts_on_safety2.pdf “A Technical Analysis of Driver Training Impacts on Safety” by ATRI, May 2008