Once upon a time, electronic batch records (EBRs) were simply a means for helping pharmaceutical companies automate away the mountains of paperwork required for maintaining production batch records. But the story didn’t end there.
Consider, for example, Baxter S.A. The diversified healthcare company is among a growing group of global manufacturers that are also using the technology to develop, implement and even enforce optimal workflows.
This new narrative began to take shape in industry in the ‘90s when the U.S. Food and Drug Administration (FDA) issued new 21 CFR Part 11 compliance standards. The regulation’s promotion of good manufacturing practices (GMPs) permitted production facilities to use EBRs to document each step in the production, packaging and handling of every batch of their products. Since then, automation vendors have shown users how automating batch records can also help them improve their workflows and efficiencies.
A case in point is Baxter’s production facility in Lessines, Belgium, where the company makes intravenously administered nutriments for patients in healthcare facilities. A new automated line there prepares solutions of glucose and amino acids and puts them into storage bags with help from Wonderware software from Schneider Electric. Not only does the software help the line conform to governmental compliance regulations in the U.S. and Europe, but it also helped Baxter double production, streamline operations and manage costs.
“EBRs are traditionally generated by MES [manufacturing execution system] or batch-management applications,” explains Michael Schwarz, product marketing manager for MES, EMI, workflow and batch software at Schneider Electric. Not only do MES and batch-management software help in managing process execution, but they also capture event execution and data records through automated means, by enforcing manual data entry, or both.
“In doing so, these software systems have significantly reduced the manual activities required to capture batch data or generate electronic reports and production documents,” says Schwarz.
At Lessines, these benefits came through Schneider Electric’s Wonderware InBatch software, which oversees batch execution, maintains equipment histories and material genealogies, and generates the necessary reports. Besides complying with the FDA’s 21 CFR Part 11 regulations, the software also conforms to the S88 batch process control standard developed by the International Society of Automation (ISA). In addition, the software relies on the World Batch Forum’s batch markup language standards to create XML (extensible markup language) documents for recipe-version histories and comparisons.
Because InBatch runs on the Wonderware System Platform, it has access to a process historian that archives production histories, compresses data, and auto-configures historical archives. Once configured, the Wonderware Historian and the vendor’s eSignature management applications ensure that the new automated line complies with the FDA’s regulations and adheres to the company’s optimal workflows. The data also gives decision-makers the historical information that they need for continuous-improvement initiatives.
Another outcome of running InBatch on the object-oriented System Platform is that Baxter was able to design a custom set of ISA S88-compliant objects representing such things as valves, sensors, weigh scales and variable-speed drives. Stored in the system’s library, these objects can be maintained from a central location and reused for future projects. Any change to an object can be linked to all instances within the system and applied only to those instances specified by the engineering staff. This ability to apply changes selectively permits the staff to validate modifications before making a global deployment.
Looking beyond the process
Documenting batch production fully, however, often requires actions beyond process automation itself. Examples include tracking quality-conformance data, compiling these records into product certificates, and documenting sign-off procedures.
“Coordinating the timely execution of quality-related activities at the end of a batch, assembling the data for documentation, and getting sign-offsare beyond the scope of the production or quality system domain,” notes Schwarz. “They still rely heavily on human collaboration.”
To oversee these interactions, automation vendors have established links to business-process and workflow-management software. “This helps to coordinate and document the activities and data flows between people and systems used in production, quality, asset management and business domains,” says Schwarz.
Schneider Electric, for example, has integrated Wonderware Skelta BPM workflow software with its Wonderware System Platform to automate human-centric workflows, as well as the processes with links to applications like enterprise resource planning (ERP) systems and SharePoint collaboration tools.
“An end-of-batch event, for example, could automatically present a sequence of quality-related activities as workflow notifications or work tasks, which in turn could initiate the required data flows or sign-off processes,” says Schwarz.
He adds that the software can replace paperwork with electronic work instructions and records even in semi-automated production environments that are still run in an operator-centric fashion. “This enables similar collaboration with quality or business functions as well as the automation of product, quality and document sign-off and approval processes,” he says.
Just by reducing manual activities in these instances, electronic workflow management can expedite the execution of EBR processes, while improving their definitions, documentation and consistency in execution.
Incorporation into ERP
Such uses of EBRs evolved alongside information technology. As capabilities multiplied and costs fell, EBRs moved well beyond the initial practice of simply scanning batch production records and storing them in an electronic repository.
Meanwhile, “as ERP-system adoption became more prevalent, EBR capabilities started to appear in stand-alone systems and in pharma-specific ERP systems,” recounts Reddy Beeram, director of product development at Edgewater Fullscope, an ERP consulting firm and Microsoft Dynamics partner. “A more recent evolution of EBR involves capturing batch production record information at the source and as the batch production progresses, from release to completion.”
Besides automating the collection and documentation of process workflows as they happen, the latest ERP systems have extended EBR capability to MESs and other classes of applications that affect batch records: product lifecycle management (PLM), laboratory information management systems (LIMS), and corrective actions and preventive actions (CAPA) software.
These developments have not been without their problems, however. Neither conventional “administrative” ERP systems nor complex niche MESs were dealing with EBRs in an efficient manner, observes Colin Masson, global industry director for manufacturing at Microsoft Business Solutions. “Many MES architectures are inflexible and don’t scale for multi-site rollouts, which is a must in today’s global, connected industry,” he says. Likewise, administrative ERP systems are not really designed for the demands imposed by the realities of manufacturing and need a fair measure of customization.
“Especially when customized, these traditional ERP and MES architectures could not cope with the constant reconfiguration and requirements to support Lean and Six Sigma initiatives,” adds Masson. They also have difficulty handling multiple manufacturing styles that have their own characteristic data models. “For example, EBRs often need to feed and provide traceability from batches through to discrete manufacturing ‘pill and pack’ operations. This is a challenge for traditional systems.”
Meeting these challenges is usually expensive and requires engineering skills that are in short supply. Because it exceeds the automation threshold that keeps many batch manufacturers from using them, they continue using manual processes based upon paper records.
To overcome the threshold for conventional MES applications and administrative ERP systems, some software vendors are offering what they call “operations ERP.” An example is Microsoft Dynamics AX 2012. “Its architecture is built to support high-fidelity, quality, compliant manufacturing,” says Masson. It is extensible and supports open web services to promote integration with mobile devices and automated data-collection devices.
The power of connectivity
This movement to make EBRs available to ERP systems and other business applications is really a manifestation of a more widespread trend—the trend everywhere toward greater connectivity. In the batch industries, the result is that more users are asking for the ability to access recipe and other information from a variety of places, reports Eric Heavin, advanced solutions and services sales consultant at Yokogawa Corp. of America.
“These users want links to workflows,” Heavin says, “so that, if they receive an order that needs to be filled by a particular day, it is sent to the DCS [distributed control system], scheduled, executed and recorded in an historical database; and results post back to the ERP system.”
Users have a raft of reasons for wanting this ability. One is to transfer data accurately throughout the organization in real time. “There is a lot of potential for error in the manual recording and transfer of data,” notes Heavin. Another reason is traceability. “Some users might keep track of it for liability reasons. They may even have to keep data for a certain number of years for their customers.”
Others simply want the ability to mine the vast amounts of historical data that is now cost-effective to collect and store for long periods of time. An example of this can be seen in a job Yokogawa did recently to help a company in the oil and gas industry improve the tracking of its products after manufacturing.
“Engineers there were trying to track the different grades going to various locations and produce a monthly reconciliation report,” recalls Heavin. “They did it by taking the pertinent data from each historical production or transaction record.”
Through an analysis, they were able to determine where the mislabeling of a product occurred and to reconcile actual product movements and transactions against the inventories in their accounting system.
In this case, the historical records were taken from a structured query language (SQL) database that had been populated over time with data collected automatically from the process. Access to a uniform set of highly granular data makes it possible for the company’s engineers to conduct various kinds of studies. Not only did they have the flexibility to track the movement of products throughout the supply and delivery chains, but they also could look at what a particular tank, for example, was doing on any day in the past.
Although the evolution of information technology has made keeping vast amounts of data for these purposes cheap and practical, Heavin warns users to store only the necessary information to avoid being overwhelmed with superfluous data. “That’s critical for presenting only pertinent data to personnel, so they can make decisions about whether a procedure or formula needs adjusting to generate improvements,” he says.