Moving Manufacturing Toward Data-Driven Safety

Summary

At many manufacturing firms today, it’s a tale of two facilities.
 
On one hand, manufacturers have significantly modernized their operations—from adding robotics and smart machines to adopting software that improves decision-making and automates workflows across the shop floor and back office. In fact, KPMG reported in its Global Manufacturing Outlook 2020 that 48% of the industrial manufacturers it surveyed have accelerated their digital transformation strategies by years.
 
On the other hand, manufacturers’ efforts to manage worker safety are still largely manual, supported by Excel spreadsheets and stacks of paper records. Lacking automation and deeper insights around workplace safety, managers too often end up reacting to incidents rather than preventing them in the first place.
 
As a result, manufacturers’ strategies to compete as trusted providers with on-time delivery on quality products and components can get derailed by workplace injuries. The human costs of an accident, alone, are high. And when a machine is involved, it will be shut down for a week or more for review and repair, putting production behind schedule and potentially leading customers to turn to other manufacturing sources.
 

Bridging the internal digital divide

The path forward is clear. In order to maximize their competitiveness, manufacturers will need to start automating their workplace safety much like they already automate other aspects of their operations. Fortunately, many companies already have at least some of the technologies in place to collect important safety data.
 
Consider the many machines that have sensors to capture data about their performance, wear, energy use, etc. Many manufacturers use real-time monitoring of this data to predict when a part is likely to start failing. Such insights can enable a company to replace the part before it starts affecting product quality or production schedules—or poses a risk of injury.
 
Beyond machines, manufacturers also have to look at other factors. Does an employee have the right training to use a particular piece of equipment safely? Is that person’s certification up to date? Are any employees working extended shifts that may lead to exhaustion and accidents? 
 
Before such data can be analyzed, it needs to be structured, normalized, and stored in a database (directly or via an application), so that it is repeatable across all existing records. Environmental, health and safety (EHS) applications are optimized for managing structured safety-related data. However, manufacturing firms that lack EHS software can get started by using their existing enterprise resource planning (ERP) and/or human resources (HR) applications to capture and analyze this information.
 
Manufacturers with no ERP, EHS or HR solution in place should consider implementing one of these applications. In our experience, doing so enables companies to start collecting data digitally and gaining important insights from their safety programs in as little as four weeks.
 
By contrast, trying to enter legacy data stored in spreadsheets, PDFs, Google docs, or paper files into a database can take months to years to complete and cost tens or hundreds of thousands of dollars. For this reason, such initiatives tend to fail.
 

Capturing meaningful insights

Once the applications are in place, manufacturers need to ensure that they are capturing the right data.
 
Generally, it’s better to collect more, detailed data, particularly as business applications start incorporating artificial intelligence (AI) and machine learning capabilities that can scan thousands of data points to find associations people are likely to miss. Trying to group information—for instance, capturing age ranges instead of each employee’s age—can limit the discovery of important insights.
 
Normalizing the data is also important. For example, manufacturers will want to know how many employees are working onsite in any given time period in order to understand the percentage of safety incidents per employee rather than simply the total number of incidents. This is especially important for manufacturing firms with seasonal peaks; here, the number of incidents alone can be misleading since there are also more workers onsite during those peaks.
 
Ideally, employees will collaborate in collecting data, such as information about hours working between breaks, training, or other safety-related factors. The key to workers’ participation is letting them use mobile phones to access web applications, which automatically normalize the data.
 
Notably, newer technologies, such as Quick Response (QR) codes, provide a way to identify individual employees via their mobile phones and effectively act as digital signatures that can be stored and tracked with other information. This can help track workers at risk if, for instance, they lack or have outdated training on particular safety measures.
 
Another opportunity to collaborate with employees is by digitizing toolbox talks or monthly safety meetings on how to safely operate equipment and machinery. A large percentage of manufacturers still deliver these safety meetings manually. By moving toolbox talks to a digital e-learning format, companies not only streamline their delivery; they also capture data about workers’ comprehension and compliance with safety requirements on the manufacturing floor. This structured data can then be analyzed to understand correlations between worker training and incidents, which can help managers to tune their safety programs.
 
By capturing detailed safety data from employees and systems across the organization, manufacturers can begin to apply the same type of analysis and reporting used for optimizing their operations to also prevent incidents, facilitate government and industry compliance, and maximize workers’ safety.