If you've read any of my recent blog or LinkedIn posts, you'll know I'm a HUGE fan of ai-driven ergonomics. It's more accurate and cost-effective than the previous systems.

BUT…some companies don't realize the full benefit. I'll outline some reasons for that below.

But first, let's define what I mean by ai-driven ergonomics – so that we're all on the same page.

Introduction to AI-Driven Ergonomics

At the fundamental level, ai-driven ergonomics usually involves a camera-enabled cell phone or tablet and some software. After you take a video of the potentially hazardous work task, upload it into the software. The software tracks body limb segments and calculates joint angles (all without any markers on the body).

Most of these software programs (and there are many of them available) base their calculations on the Rapid Entire Body Assessment (REBA) or the Rapid Upper Limb Assessment (RULA) formulas for calculating an overall risk score for the task. These two assessments are the gold standard in ergonomics for calculating an overall risk score.

After the analysis (typically lasting less than 5 minutes), you will have an overall risk score for the task and individual risk scores for various limb segments. The limb segment scores help you focus your hazard reduction efforts.

The final step is synthesizing the software data, understanding the task being performed, and then making recommendations for hazard reduction. This last step is where expertise in ergonomic risks and countermeasures is critical…something the software cannot do for you.

How AI-Driven Ergonomics Work

ErgoScience has over 30 years of expertise in helping enterprise-level organizations create and strengthen their safety cultures using our proven research-based injury prevention programs. Some of these programs include pre-hire post-offer physical abilities testing; job demands analysis, and AI-powered computer vision ergonomic risk assessment and training. We understand the importance of a strong safety culture and have helped numerous organizations significantly reduce incident rates and operating costs.

Why AI-Driven Ergonomics Doesn't Create Maximum Benefit for Some Organizations

Sounds simple. What can go wrong?

In our experience, there are three main reasons why ai-driven ergonomics doesn't create maximum benefit for some organizations.

Lack of Expertise

It's one thing to upload the video to the software and calculate a risk score. With most programs, only a few minor manual inputs are necessary. But once you get the risk score, what recommendations will you make?

• Will you try to train workers in a different procedure or different body positioning? If so, what is the most effective way to teach them?
• Will you recommend different positioning of materials or supplies? A new racking system? Gravity flow or motorized conveyors?
• An overhead hoist?
• A two-person lift?
• Will you suggest a job rotation? Do you have enough job variability for that? Rotating someone from one shoulder-intensive job to another shoulder-intensive job won't help.
• More proactive equipment maintenance?

The great thing about ai-driven ergonomics is that you can make before and after assessments. If you cut the risk score in half, that's great…but will it be sustainable?

If all of your recommendations are contingent upon employees following a new procedure or using new equipment, will they be able to follow the new processes? If not, your recommendations won't be practical.

Knowing which recommendations work in a given situation dramatically depends on understanding the corporate safety culture. If an organization truly puts safety first, administrative or training solutions might be more effective. But your training and administrative solutions might not make much difference if productivity trumps everything else- even safety.

Some organizations expect their safety people or front-line supervisors or managers to become experts in ergonomics with a brief course of instruction. Others claim their insurance carriers have safety experts that will do their assessments for "free" (not really because the cost is built into their premium cost).

In our experience, these individuals are typically trained in some basic ergonomic concepts in a short course and equipped with the software. Still, they lack the proper expertise that comes with professionals who have extended training in human movement and biomechanics, such as physical and occupational therapists and ergonomists.

Some organizations have found success using a combination of internal associates for ai-augmented data collection, supplemented with consultants who review the videos and reports and collaborate with internal staff on recommendations.

Lack of Budget & C-Suite Support For Ergonomic Changes

AI-driven ergonomic assessments are great. But if there's no budget for training or making ergonomic changes, why do the evaluations in the first place? You need the resources to fix the problem.

Perhaps you'll use the data for the upcoming year's budget rationale or justification?

At any rate, the C-Suite has to be behind the project. And getting C-Suite buy-in can be challenging. They must see that the project will have a positive return on investment (ROI). Research at the Liberty Mutual Research Institute for Safety shows that $3-$6 is saved for every $ 1 invested in safety.

On average, employees who work in pain experience a 10% reduction in productivity if they are working in pain. If an employee earns $50,000/year, as much as $5,000 in salary per year might be lost to discomfort. Not to mention the lost revenue to decreased productivity.

How much more productive could your employees be if you could increase their comfort? How many individuals in your organization are working with pain or discomfort day in and day out?

When indirect and direct costs are considered, one lost-time musculoskeletal injury costs approximately $60,000-$80,000. How many of those injuries does your organization experience each year?

These figures and an estimated project cost can help determine the project's ROI. And targeting the jobs producing the most injuries can help you narrow the project scope so that your organization realizes the most benefit for the least expense.

Ensuring you can align your ergonomic project with one of your organization's strategic initiatives also helps justify the project in the eyes of your executives.

Employees Don't Embrace the Changes

One of the best ways to ensure that your ergonomic assessments and countermeasures won't lead to meaningful change is to fail to ask the employees for input. Employees who do the job day after day have ideas about making the job easier and more efficient. And your ergonomic changes need to be compatible with their ideas.

Every time I've walked into an organization, I'd had a nickel for every time I saw heavy material handling that equipment could minimize. And when I investigated further, I found they already had that equipment. But it was sitting in the corner collecting dust.

Why? Because no one consulted the employees, and for one reason or another, the equipment slowed them down or wasn't effective in reducing the load. If only they'd been consulted…

When videotaping hazardous tasks in the field, it's an excellent time to ask employees what they think. If you tell them you're trying to make their jobs easier and ask for their opinions, they'll likely share valuable information that can guide your recommendations. And their cooperation will be much better when it comes to implementation.

When it comes to ergonomic training, ai-assisted ergonomic assessments can show the difference in hazard scores before and after the recommendations are followed. The dramatic decrease in hazard scores motivates employees to adopt the new, more ergonomic work practice. As the old saying goes…a picture is worth a thousand words!

Success with AI-Assisted Ergonomics

Success with ai-assisted ergonomics hinges on getting three things right:

1. The right level of expertise. Going it alone might not yield the result you desire.
2. The proper support from your executives. Get them on board through ROI and strategic alignment.
3. The excellent cooperation from your employees. Ask for their input.

If you would like more information on ai-augmented ergonomic analysis and recommendations, you can contact ErgoScience. [INSERT LINK to Meeting form?]

What is ComputerVision AI?

ComputerVision AI is a broad category of technology that applies to everything from facility management to targeted ergonomic assessments. For this blog post, I will refer to ComputerVision AI for targeted and detailed ergonomic assessments. In this niche application, the software builds complex 3D models of the human body moving through space from standard videos – videos recorded through a mobile app or uploaded on web-based software. With this information, the system completes Industry-standard ergonomic assessments that quantify risks with a standard methodology that prioritizes interventions to prevent workplace injury.

After a video is uploaded, the system automatically determines the hazardous postures in a job, the frequency of movements, and the duration of postures held. Then the AI recommends which postures and parts of the body the Industrial Athlete should adjust to mitigate the risk of injury.

Benefits of ComputerVision AI. There are many benefits to using ComputerVision AI in the field of ergonomics:

• The user obtains more objective ergonomic hazard analysis data
• The hazard data covers the entire task, not just one brief snapshot of the task
• The resulting data indicates the limb segments that are creating most of the risk
• The program also predicts how much the industrial athlete can mitigate risk by addressing each limb segment.

ComputerVision AI Helps with OSHA Compliance.

Another huge side benefit of using ComputerVision AI is that it can help organizations remain in compliance with the requirements of OSHA 29 CFR 1910.269(a)(2)(iv). Paragraph 1910.269(a)(2)(iv) states the following: "The employer shall determine, through regular supervision and thorough inspections conducted on at least an annual basis, that each employee is complying with the safety-related work practices required by this section."

It's important to note that OSHA changed the word "and" in the paragraph above from "or" in 2014. In other words, employers can no longer claim compliance with OSHA 29 CFR 1910.269(a)(2)(iv) with just regular supervision. This change is significant. Conducting inspections of field personnel has become a regulatory compliance issue. Supervisors should perform a second observation after the industrial athletes implement any corrective action to ensure they follow through on the AI's recommendations.

Ergonomics is essential in any field inspection where workers handle materials, maintain their positions for extended periods, or perform repetitive movements. Of course, ergonomists can conduct field inspections without ComputerVision AI. BUT… they're more subjective, cumbersome to complete, and often employees are disengaged with the process. In addition, supervisors and managers are hesitant to do them because ergonomics isn't their expertise. ComputerVision AI does a lot of the work for the management team and provides them with the expertise they need.

ComputerVision AI Focuses on Leading Indicators.

OSHA requires employers to record their injury and illness statistics. But these statistics are lagging indicators. They often have little to no impact on the safety practices of an organization and the prevention of future injuries.

In contrast, regular field inspections provide leading indicators because they show whether employees apply safety principles in the field. In the case of ergonomics, field inspections show whether employees are lifting and carrying materials safely. They also identify situations where employees reach beyond the safety zone to obtain materials and supplies and get into unnecessarily awkward positions. Or when slightly changing work practices would minimize unnecessary risk. AI can prevent many, if not most, injuries by identifying these ergonomic hazards and evolving work practices, workstation setup, or using tools and equipment differently. 

ComputerVision AI helps managerial staff and safety managers conduct these inspections with confidence that they are basing their recommendations on objective data and not just subjective opinions.

ComputerVision AI engages frontline employees in the field inspection process.

Field inspections are not always well received by employees. But ComputerVision AI changes employee attitudes even within one inspection session.

How? Employees are curious about the technology. Evaluators can show them a previous example with their cell phones or tablets. Then employees start to wonder what they would look like if they improved their technique. What would their scores be? And at that point, their engagement pivots. They are more eager to be videotaped. In addition, the software can block their faces for privacy concerns. Once they receive feedback, are videotaped again, and see the dramatic changes in their scores, they are motivated to change.
When the session concludes with praise for a job done well and safely, observer-employee-employer relationships grow.

Reassessment with ComputerVision AI.

ComputerVision AI helps to determine the effectiveness of training, ergonomic modifications to tools and equipment, and workstation modifications. Performing observations with ComputerVision AI helps companies track data and analyze trends to determine future ergonomic-related needs, such as training, modifying workstations, and purchasing new PPE, tools, or equipment.  

Before You Begin Using ComputerVision AI

Starting a ComputerVision AI ergonomic inspection program at your company requires purchasing software or hiring ergonomic consultants specializing in ComputerVision AI inspections and training. Naturally, you'll have to get buy-in from leadership which means explaining the importance of meeting OSHA's requirement for field inspections, explaining the role of ComputerVision AI ergonomics in injury prevention, and, last but not least, presenting a projected return on investment.

OSHA and the National Council on Compensation Insurance's (NCCI) Workers Compensation Statistical Plan Database have statistics that can support your projections. Since the average direct and indirect costs of a lost time workplace injury are over $80K, it's often easy to justify these programs.

Once they sign on to the ComputerVision AI program, leadership should regularly participate in field observations. Regular observations will support field personnel's acceptance of the inspection program and help to strengthen the safety culture of the organization.

Corrective action policies are essential to consider before developing the inspection program. If you have a policy that reprimands employees based on the inspections or is otherwise negatively focused, it will be tough to get buy-in regardless of the technology used. You may need to negotiate a new policy or an addendum to the existing approach to allow for inspections and coaching without reprimand.

Conducting Inspections with ComputerVision AI

Safety management personnel, operations supervisors, and managers can conduct the ComputerVision AI ergonomic inspection program. ErgoScience experience has shown us that even though the technology is extremely user-friendly, those implementing ComputerVision AI will likely need some training in basic ergonomic principles, use of the technology, and field practice under supervision to maximize the success of the program. In addition, we provide a hotline for questions regarding the use of the technology or the appropriateness of recommendations.

One of the benefits of having supervisors or managers perform field inspections is that they can connect with frontline personnel, hear any concerns, and give positive feedback for correcting unsafe work practices. An advantage of having safety personnel conduct the field inspections is that they typically have a little more experience and knowledge, depending on their backgrounds, in the field of ergonomics.

For managers and supervisors to be effective in performing the inspections, they need training in the following:

• The basic principles of ergonomics and safe work practices in manual materials handling
• The use of the ComputerVision AI software and interpretation of the scores.
• Coaching, explaining the software outputs to employees, and providing corrective feedback.
• Providing sincere, positive feedback for improving safe work practices.
• How to address concerns regarding equipment or workstation setup.

When a non-ergonomic infraction occurs during an ergonomic inspection, it presents an opportunity to find out why it happened. For example, perhaps a worker isn't wearing the appropriate PPE because of poor fit and discomfort. The solution would be to provide the worker with better-fitting PPE. Managers and supervisors should address major life-threatening safety infractions more firmly.

Scoring and Tracking

Once the videotaping is complete and uploaded into the software, the software calculates an overall task hazard score. It indicates which body segments contribute the most to the overall hazard score. The software also notes the percent improvement in the overall hazard score expected if the employee changes that body or limb segment position. Management can share the scores with the employee, and together the manager and frontline worker can problem-solve the best work practices to address the hazards. It might make the most sense for the organization to change the workstation setup or positioning of materials altogether to further aid in improving the scores.

Managers/supervisors need to recognize good work practices whenever possible and then offer feedback about what needs improvement to achieve a better score in the future.

After the audit is complete and supervisors share the results with employees; management can add notes to the reports stating what advice the employee received and attach another video of the worker using the recommended work practices. The safety team can use this method to provide a comparison report documenting change from pre to post-feedback.

Summary

OSHA requires field inspections, and ComputerVision AI can help make the process easier, more efficient, and more objective. Managers and supervisors are more willing to conduct the audits. Frontline workers get significantly more engaged than with standard ergonomic assessments and feedback. ComputerVision is a valuable tool that can help a company determine if evaluations and feedback have been effective. Maintaining documentation of these inspections keeps organizations in compliance with OSHA.

Ergonomics In The Workplace

As long as physically demanding jobs exist, there will be efforts to make them safer and less risky for both the employee and the employer. According to the National Safety Council, work-related injuries cost US companies nearly 164 billion dollars in 2020; that's almost half a billion dollars per day!

While the cost is increasing, work-related injuries have always been an unfortunate part of the landscape of physically demanding jobs. Until the robots completely take over, they will remain so. This fact, combined with data that shows our workforce is getting older and, therefore, more susceptible to musculoskeletal injuries, pushes us to identify new technologies to reduce, or better yet prevent, these incidents from occurring.

Injury Prevention

As injury prevention experts, we're constantly searching for ways to recognize the hazards of physically demanding jobs so we can attempt to decrease the risk of injury. We're all familiar with the hierarchy of controls when it comes to job safety:

1. Elimination – Physically removing the hazard
2. Substitution – Replacing the hazard
3. Engineering Controls – Isolating workers from the hazard
4. Administrative Controls – Changing the way people work
5. PPE – Providing workers with proper Personal Protective Equipment

These are listed in order of most effective (removing the risk) to least effective (providing PPE).

Removing all hazardous activities from every job would be fantastic, but that's unrealistic. Some jobs are inherently risky; hopefully, employers will do their best to make them as safe as possible.

What happens when a job, or one of the tasks required to perform that job, can't be eliminated?

Preventing Injuries With Wearable Sensors & Ergonomics

Ergonomists have historically been called upon to observe job-related tasks, identify the risky behaviors involved with performing the task, and give suggestions on decreasing the risk. There are limitations, however. It is challenging to observe every aspect of every job; some hazardous tasks infrequently happen under normal circumstances, and some only occur in emergencies. An action that isn't observed can't be improved or removed.

What if there was a way to "observe" an employee and collect biomechanical data on movements and postures the whole time they were working? It would cost a fortune to have an ergonomist on site all day following each worker, and it would likely cause a massive disruption to production.

Thankfully, technology allows us to do just that without physically following an employee's every move.

Wearable sensors have been developed that track joint angles, postures, and repetitions of movements, not to mention heart rate, location in a facility, proximity to heavy equipment, temperature, light, humidity, decibels… Everything informs us about preventing injuries without needing a human following the worker and with much greater detail than an ergonomist could provide.

For example, an ergonomist can observe a worker performing a lifting task and approximate the angle of bend for the trunk, shoulders, and knees. They could count the number of repetitions that worker performs that specific task while observing and then extrapolate to come up with an approximate total for a whole day.

Put a wearable sensor on that same employee, and you'll collect exact angles for the involved joints and the precise number of repetitions of each task. In addition, you can identify trends like days of the week, or even times during each day, when employees are exerting themselves maximally, which makes industrial athletes more likely to get injured. Also, you can recognize specific periods when biomechanics begin to break down (do behaviors become riskier later in the week when fatigue may be more of a factor?).

You can use the data to identify workers with good body mechanics and use that information to inform other workers on proper techniques.

You can collect data specific to each individual to know precisely what coaching method will effectively reduce the risk for that employee.

Changing Behavior With Wearable Sensors

Another benefit of wearable sensors is that often you don't even need to wait for a "coaching" opportunity face-to-face. Many sensors can provide immediate haptic feedback (a light but noticeable vibration) to indicate when a risky posture or activity is taking place. This type of immediate feedback can be much more effective at changing behavior than traditional one-off coaching that occurs randomly after the behavior occurs.

Once behaviors have been addressed and changed, sensors allow employers to track those changes to see if they stick and revisit them when necessary.

Contact ErgoScience Today!

So, how do you implement this technology? At ErgoScience, we've partnered with some of the world's leading wearable sensor companies. Not only can we help you get the technology into your facility, but we also have the expertise in ergonomics and biomechanics to help you make sense of the data to bring your injuries down and give you the greatest possible return on investment.

Contact us today to find out how easily you can bring your ergonomics program into the 21^st century!