Most nuclear plant trips are caused by human error . Many mistakes occur during preventative maintenance due to not following or maintaining procedures. Unfortunately, most plants today use static paper or PDF-based procedures, which are prone to human errors and inefficiencies. Recent advances in smart procedures are addressing these problems. With capabilities such as conditional logic, data entry validation, automated calculations, barcode scanning, video references, qualification checks, push notifications, and real-time job progress tracking, smart procedures save plants money by making the craft safer and more efficient in the field.
There are many risks associated with human error and inefficiencies in the nuclear industry. Errors can impact personal safety and station availability and have a direct effect on fiscal viability. Anything that competes for worker focus has the potential to impact decision-making, and these distractions often result in errors or delays. Even simple tasks like branching to another section of a procedure or performing a calculation in the field can create opportunities for human error. Errors and delays ultimately increase the cost of plant maintenance and operations. Reducing the likelihood of these errors and minimizing job delays are essential to worker productivity and plant viability.
Over the past 20 years, procedural guidance has helped make nuclear energy safer, more efficient, and more reliable. Along the way, however, procedures became longer and more complex to address variations in worker skill, plant configuration, and field conditions. As a result, today’s workers are provided with lengthy paper or PDF procedures, which are difficult to navigate, prone to human error, and cumbersome to use in the field. Although moving procedures to a PDF format has reduced the physical volume of work packages, it has not resolved the issues inherent in static, paper-based legacy systems . The industry’s strategic plan, “Delivering the Nuclear Promise®,” has championed the elimination of inefficiencies while maintaining or improving quality. Yet, to date, planning and work execution areas have not seen any significant efficiency gains or quality improvements. In fact, the steady increases in procedure and process complexity over the last two decades have had a negative effect on worker productivity.
With recent technology advances in alternative energy sources, like natural gas power generation and renewable energy, there is continued pressure on the nuclear industry to be more competitive in the energy marketplace.
Complex work instruction documentation forces workers to search for relevant sections on the job. Many of today’s procedures originated as simple steps provided by manufacturers as part of their technical manuals. Over the years, those steps have been buried in additional content, including revision history, specifications, safety reminders, references, and regulatory guidance. All this additional information has increased procedure volume and complexity. The industry needs a solution that improves work execution efficiency, reduces opportunities for errors, and makes it easy to maintain a continually growing knowledge base.
Work delays have the largest impact on worker productivity and are often the result of ambiguous procedures, poor communication, or inadequate planning. In addition to wasting resources, work delays interrupt work pace and introduce distractions that can cause errors. Work hand-offs and need-times for support providers (anyone outside of the primary execution team with at least one action to complete in the work sequence) move as work pace changes. Keeping everyone in sync with actual work progress to avoid delays is difficult. Supervisors and foremen need to be able to monitor the work progress of their crews in real time so that they can proactively address in-field delays and ensure timely job completion.
When workers are in the field, many elements compete for their attention: safety hazards, announcements, rotating machinery, and co-workers. Without the worker’s full attention, when presented with even simple calculations or decisions, mistakes are often made. A mistake on a calculation or assessment can result in performing the wrong steps, not notifying authorities of out-of-spec conditions, or returning equipment to service that may not be reliable. Even something as seemingly simple as not verifying the correct equipment prior to starting work has resulted in reactor trips, equipment damage, and injury .
Work order duration and worker assignment require knowledge about the duration of a scope of work, skill requirements, and the craft’s qualifications. As experienced schedulers and work management resources leave the workforce, the knowledge they’ve acquired over the years leaves with them. Work management systems are much less effective than departing workers at accurately capturing or retaining this information. Although some systems allow manual entry of scheduling and resource data, the information is not easy to manage and is often not trusted. More importantly, this data is crucial to plant maintenance scheduling and resource planning.
Since planners have been planning work on plant equipment for more than three decades, most scopes of work have already been planned. The problem is that plans are not consistently saved or cataloged for convenient reuse. Since individual planner assumptions about resources, support, and the availability of reference materials differ, plans tend to differ as well--even for the same scope of work. Different instructions to accomplish the same work can create confusion that results in mistakes.
At a fleet level, the problem is magnified. Since plant equipment is often the same across many locations, repair plans and maintenance routines could be identical, but often are not. Three stations in a fleet with the same equipment and vendor manuals often have different procedures for doing the same work. In addition to wasting time, these discrepancies introduce the risk of worker confusion. When workers shared between stations use different instructions to accomplish the same work, it’s easy for them to get confused and make mistakes.
First generation Electronic Work Package (eWP) solutions prioritized getting legacy content into workers’ hands via mobile devices by rendering paper documents into PDFs, but these solutions failed to assess and address the planning and work process as a whole. In fact, they created new problems and increased the complexity and duration of the work process. Presenting legacy documents styled for an 8.5” x 11” sheet of paper on a device that is marginally 60% of the size of a sheet of paper creates usability issues. In response to this issue, first-generation eWP users tend to require large, heavy mobile devices in order to legibly view a PDF on a tablet.
Workers in the field often have difficulty navigating PDF documents and entering data on mobile devices. They still have to comb through the same volume of content as they did with paper procedures in order to find a reference or applicable set of instructions. Data entered on a marked-up PDF is an annotation that is not retrievable or accessible as discrete data. Even with smart PDF overlays, entered data is not easily retrievable. Notes and comments made in the margins are often lost unless, during review, someone takes the time to transpose them into formal actions for resolution. Extracting the QA record content from an annotated PDF also requires significant investment in administrative resources. The work package must be reviewed to identify QA content so it can be assembled into a QA record for submittal. This process has proven to be more demanding than the legacy paper process, and as a result, the entire work package is often submitted as the QA record, which creates waste and inefficiencies.
Most importantly, first-generation solutions barely harness the technological power of mobility or the potential of connected, collaborative work environments. Automated Work Packages (AWP) are the logical evolution of eWPs. They incorporate the advanced automation technologies and innovations of the future and address the work process deficiencies not resolved by first-generation eWP solutions available today .
Lean Power was created to support and enhance the entire work management process, with an emphasis on reducing human error, increasing worker productivity, and providing real-time access to job tracking. Equipment attributes, in-field conditions, and worker progress adaptively drive the work process while increasing efficiency and reducing human error. Lean Power is a software application that can be used to produce smart procedures fundamental to AWPs: electronic, intelligent, and interactive work packages . The solution is comprised of four functional areas: editing, publishing, work execution, and analytics .
Figure 2: Work order completion time by execution method Editing
Procedure writers can use a web-based editor to create, edit, and manage smart procedures. The editor is a simple, intuitive Microsoft Word-like interface where users can enrich their procedures with smart logic, such as data entry steps, condition-based steps, calculations, complex tables, signatures, and more.
Lean Power’s smart procedures also support a rich set of references, including PDFs, images, videos, drawings, and references to other digital procedures. The editor also contains a parser that easily converts legacy Microsoft Word document procedures into Lean Power smart procedures.
In addition to creating smart procedures, planners can use the editor to manage work plans. With the editor, planners can organize discreet instructions, procedures, and procedure sections into a sequence of work, saving them as a work plan. Work plans are extensible, reusable, and can be applied to other equipment. Planner-specific interfaces are provided to help the planner find and bind work plans to on-demand work orders. Planners can also bind work plans to repetitive tasks, so that each time a task is generated, the correct work plan is automatically assigned.
Figure 3: A screenshot of the Lean Power editor Publishing
Lean Power can publish smart procedures and work plans in a variety of methods for execution in the field. The publisher output is contextual, so content can be customized to a specific piece of equipment, equipment type, or environment. Publishing can create instructions for a single scope of work on a single piece of equipment, a single scope of work on several pieces of equipment, multiple work scopes on a single piece of equipment, or multiple work scopes on multiple pieces of equipment.
Publishing is performed on-demand, so when a worker requests a work document, the publisher creates an instruction set using only the latest procedure revisions. Procedures and work plans can be published to a mobile device, web browser, or PDF file. There is no limit on the number of mobile devices to which a given work scope can be published. The web version can be used by people with browser access who are not in the field. A PDF output can be ported over to a first generation solution or printed and completed as a hard copy.
The execution application is the interface for executing work in the field. Hardware- and operating system- agnostic, the application does not require a wireless connection. By enabling collaboration, employees at different locations can work on the same plan or procedure. All user input is recorded and made available immediately when the user connects to the Lean Power server. During execution, place-keeping, discipline, and qualification requirements are enforced. This means that you can configure certain procedure steps to only be completed by qualified users. Instruction tasks and data entry are interactive and provide the worker with visual confirmation as work tasks are completed.
Data entry values can be validated and used as inputs into conditional logic that directs the worker to the appropriate next action. Data entries can also be used as input into calculations where the output of the calculation can be evaluated against limits and used as an input into conditional statements.
Conditional content and steps that have been evaluated as not applicable during execution are still visible to the worker, but these steps are struck through, marked as “N/A,” and cannot be performed. That way, workers can’t accidentally perform the wrong steps.
Equipment, concurrent, and independent verification steps are supported and can be integrated into procedures and work plans. Verification steps can also be used to confirm that materials, tools, and locations are correct.
Work pace is a great example of a powerful, proactive indicator that can be used to reduce waste in your work plan. When the work is executed, a detailed step-by-step history of each activity is recorded and made available for analysis. As execution history accumulates, Lean Power can calculate historical execution work pace and compare it to current work progress. With this analysis, it is easy to determine whether work is behind schedule, ahead of schedule, or on-track. Work that is behind schedule can be proactively supported to get it back on track. This single metric has the potential to dramatically improve work week and refueling outage efficiency.
As tasks are completed, pre-defined notifications can also be sent to supervisors and support personnel to provide work progress or to prompt a support action.
Figure 4: Screenshot of Lean Power execution mobile app Analytics and Enterprise Integration
Lean Power can integrate with core enterprise solutions like asset management, document management, and project scheduling systems to produce a seamless, coordinated work management solution. Lean Power can also integrate with learning management, lockout-tagout (LOTO) solutions, and other enterprise tools to further extend Lean Power’s functionality and analytic capability.
During planning and work execution, a tremendous amount of data is captured. This information combined with enterprise data can be transformed into actionable information using Lean Power analytics. Temporal, hierarchical, and multidimensional data visualizations are combined to create dashboards aimed at specific roles ranging from field worker to executive.
Figure 5: Screenshot of Lean Power analytics
Throughout the development of Lean Power, we constantly questioned how to improve the work process while minimizing risk. Let’s review how we’ve addressed some of the most common questions.
How can we make instructions clear and concise without oversimplifying the content?
Digital procedures retain all the original procedure content. The difference is how the content is presented to the user. Emphasis is placed on clearly communicating the action expected from the user. Each procedure step has a step type where user action is designed to be intuitive, user-friendly, and simple. Content intended as reference material is available in line and accessible from the reference library. Steps and content not applicable to the equipment or work being performed are not presented to the user during execution; instead, they are accessible in the reference library. The resulting work sequence is clean and concise. Place-keeping is integrated with step completion and is enforced with intuitive visual cues. Ordered and unordered step execution is also supported. The user is presented with a continuous set of steps instead of a series of procedure references, and as a result, the user no longer needs to move between documents to execute work.
How do we reduce delays and give supervisors insight into work execution?
Work progress is known down to the current step unless work is being performed offline. In cases where a scope of work has been performed in the past, work duration and work pace are known and can be used by supervisors to proactively engage with workers to prevent delays. Delays can also be caused when work progress is unknown and expected actions are not proactively communicated to support providers. Notifications can be placed in procedures and work plans at points where a support provider should be informed of an upcoming task. If the work is executed online, the notification is launched when work progress reaches that point. If the work is executed offline, the notification becomes a task that instructs the worker to notify the support provider using other methods. Either way, the support provider should get the notification, and this action will be documented. This simple change has a profoundly positive impact on ad-hoc in-field work coordination and waste reduction.
How can we reduce the number of in-field decisions made by workers and make the remaining ones more objective?
Most in-field decisions are related to assessing some condition and then choosing a course of action based on that assessment. A solid human factors approach is to ask the user to record the value related to the condition being evaluated. Once the value is recorded, a predetermined course of action is presented to the user by the device. To simplify data entry, radio buttons or pick lists can also be used. Conditions with a binary choice, like “good” or “bad,” are more subjective than a set of descriptive choices, such as “no scratches,” “minor scratches,” “cross-face scratches,” “pitting,” and “corrosion.” In either scenario, once data is entered, the device performs the evaluation and designates the correct course of action.
How can we leverage work history and execution data to make scheduling and planning less subjective and more efficient?
A common approach plaguing the industry is the scheduling of work durations in half-shift increments. The main reason is that schedulers have limited means of determining actual work durations. Work duration is captured each time a scope of work is executed. Since execution data is detailed (including the identities of work performers, their teams, their locations, and so forth), schedulers can leverage execution analytics for a given scope of work and project a highly accurate duration. They can also leverage that same data to optimize personnel assignment and work plans during critical work times, like outages. At that point, planning becomes heavily data-driven. Subjective planning only occurs when the work scope is not understood, the workers are inexperienced, or the planner lacks experience planning for the given scope of work. Ad-hoc feedback provides a communication conduit between the worker and the planner. Planning omissions or errors can be caught during work preparation or execution. In either case, worker comments are available at the work step level to promote prompt feedback to the planner. With work plans, changes can be made rapidly and are available to the worker as soon as the package is acquired. Changes to procedures can also be made rapidly and are available as soon as the users’ revision, review, and approval protocols are satisfied.
How do we preserve planning techniques and knowledge so that repair and maintenance plans are consistent and reusable?
This is accomplished by making the creation of a work plan, finding an existing work plan, and applying existing work plans to new work easy to perform. Each time a work plan is created, the creator has an opportunity to bind metadata to the work plan. Metadata entered by the creator includes applicable equipment type(s), a simple work description, outage indicator, mode restraints, etc. The more descriptive, the easier to find and reuse a work plan. Metadata is also updated each time a work plan or procedure is applied to a piece of equipment or work scope. Locating a procedure or work plan applicable for a scope of work on a piece of equipment is as simple as finding content on Google. Reusability of work plans correlates directly to the ease of locating applicable work plans versus creating new ones; you cannot reuse something you cannot locate. Additionally, when a new work plan is created, the planner can search for and copy an existing work plan as opposed to creating one from scratch.
How do we move forward with smart procedures?
Smart procedure implementation is intended to be an incremental and organic process, to be performed alongside existing solutions. We recommend initially using Lean Power for work scopes that are performed frequently or have demanding data input requirements. Conversion of legacy procedures to basic digital procedures can be accomplished quickly using the parser and a small time investment to enrich procedures with smart logic. Additional time may be required to convert procedures with complex tables and large data entry demands, while little time is required to convert procedures that are primarily a sequence of tasks. The motivations to move forward should be error reduction, improved worker productivity, and gains in process efficiency.
The team at Lean Power understood early on that the focus of any AWP solution had to be heavily biased toward helping the worker be more successful. Two large measures of worker success are increased productivity and reduced errors. Lean Power leverages mobile technology, distributed tasking, and collaborative work environments to accomplish these outcomes. An intuitive interface that creates a positive user experience is essential for a solution to be accepted by workers. Finally, the solution must make a positive and measurable difference to the cost of doing business.
Lean Power is the culmination of coupling detailed knowledge and experience in maintenance and work management with cutting-edge design and development. Lean Power ushers in the new generation of AWPs by delivering a holistic, user-centric solution that embraces digital procedure creation, editing, publishing, work execution, and analytics.
This work would not have been possible without the support of Devbridge Group. The authors gratefully acknowledge Martin Stasaitis for recognizing the potential of automated work packages two years ago and for his unwavering commitment to the Lean Power initiative and the Lean Power team. The development of this white paper benefited significantly from the input and support provided by our outstanding team of reviewers.
Work delays have the largest impact on worker productivity and are often the result of ambiguous procedures, poor communication, or inadequate planning. Unfortunately, most plants today use static paper or PDF-based procedures, which are prone to human errors and inefficiencies. Recent advances in smart procedures are addressing these problems.