Mergers and acquisitions have long been a staple of the life sciences industry — a reality that has resulted in disparate information technology (IT) environments within and across often far-flung enterprises. The situation is particularly prevalent in manufacturing operations because a pharmaceutical or biotechnology organization may run different management systems in nearly every separate facility (especially when contract manufacturers are involved). These “siloed” information environments are hindering efforts to evaluate and improve quality and operational efficiency at a time when quality and efficiency are most important to life-science executives.

One way to solve the problem is to implement a single IT application across all manufacturing plants. But that is risky, expensive, and time consuming — if at all feasible given good manufacturing practice (GMP) and validation requirements as well as business pressures. Biomanufacturers, therefore, seek strategies and tools to help them make the most of their existing environments while enabling quality and efficiency globally. The answer could be enterprise manufacturing intelligence.

The Integration Factor

Life-science organizations facing rising costs, declining pipelines, and margin erosion are increasingly focusing on improving manufacturing efficiency and reducing both cost and waste without compromising quality. To achieve those goals, companies must aggregate and rapidly analyze manufacturing data across devices, equipment, production lines, and plants — sometimes across borders.

Improved manufacturing intelligence offers significant benefits. AMR Research (www.amrresearch.com) estimates that such initiatives can yield production efficiency improvements of ≤25% and improve cycle times ≤20% by providing manufacturers with the information they need to make informed decisions about scheduling, production, and scrap (1). For example, pharmaceutical companies implementing enterprise manufacturing intelligence initiatives can determine the best way to efficiently produce high-quality drugs in the shortest possible time with existing and potential resources. They gain the ability to accurately assess options such as running double shifts, increasing or decreasing outsourcing, or expanding manufacturing capabilities.

Enterprise manufacturing intelligence can also help organizations increase their asset use by ≤10% annually because they can tell when a machine may be slowing down and thus schedule preventive maintenance accordingly to prevent outages. Similarly, expanding intelligence into equipment availability enables companies to optimize the use of their production machinery through improved agility in scheduling and rerouting. These capabilities are critical when manufacturers want to optimize their investment in production facilities and tight working capital.

By integrating manufacturing data, life-science organizations can also eliminate discrepancies between manual and machine data. Thus, preventive maintenance can be performed on the basis of machine “actuals” versus “best guesses,” averting unplanned downtimes and ensuring optimal schedules for equipment maintenance. That in turn reduces on-hold and other production-delay issues.

Quality has always been under the microscope in bioindustry. Integrated manufacturing data provide critical support for quality initiatives such as lean manufacturing, six-sigma, and quality by design (QbD) that require real-time and enterprise-wide information. Business intelligence based on quality data will successfully identify waste and outliers as well as variability and inconsistencies.

Finally, biopharmaceutical companies that launch enterprise manufacturing intelligence initiatives stand to gain a competitive edge through their ability to share quality information with trading partners. Integrated manufacturing data enable companies to connect and collaborate more efficiently with a growing list of manufacturing partners. Integration helps them capture and analyze information more effectively to improve quality as well as comply with regulatory requirements.

Solving the Puzzle

Most organizations are clear about their enterprise manufacturing intelligence objectives. The challenge lies in achieving expanded visibility in a disparate and resource-constrained environment. Today, many life-science organizations still rely on spreadsheet-based attempts to gain greater manufacturing intelligence. But the process has several drawbacks: It is resource intensive (and error prone), and the end results do not yield real-time information needed to drive the level of agility required in today’s market.

Manufacturers have several options for automating visibility into their manufacturing operations at an enterprise level. Implementing a unified manufacturing execution system (MES) across an entire enterprise could provide end-to-end visibility and real-time information required to improve manufacturing efficiency and assist with quality initiatives. But cost, time, and validation complexities all come into play with this approach — making it out of reach for most companies in today’s economy.

Life-science organizations might also consider building a network of point-to-point integrations into a centralized analytics environment. But this approach is resource intensive to build and maintain. It also adds complexity to a company’s IT environment and cannot be deployed quickly, ultimately increasing time to value and total cost of IT ownership.

A third approach involves creating a manufacturing data hub and common extraction layer that helps a company collect information from various plant systems. A reporting tool sits on top of the infrastructure. This approach provides the visibility required for precise management of a manufacturing environment, can be deployed relatively rapidly, and offers the flexibility to quickly incorporate data from new systems.

Adopting Best Practices

As life-science organizations consider approaches to enterprise manufacturing intelligence, several best practices can guide them to a successful initiative. These include understanding the scope of data integration, sticking with standards, building in system flexibility, accelerating payback, offering “something for everyone,” and providing actionable information.

Understand the Scope of Data Integration: An enterprise manufacturing intelligence environment must capture and analyze real-time, granular information from all types of systems, including programmable logic controllers (PLCs), supervisory control and data acquisition (SCADA) systems, and distributed control systems (DCS) on plant floors. Access to such information helps a company build quality into its processes, reacting to exceptions and quality deviations as they occur, rather than after quality control testing.

In addition, pharmaceutical manufacturers should ensure that their manufacturing intelligence environments can integrate data from third-party sources such as contract manufacturers. This capability is important because even when they outsource, pharmaceutical manufacturers remain responsible for quality data involved in their own products and processes.

Stick with Standards: The most flexible enterprise manufacturing intelligence environments are built on standards that companies can use to leverage prebuilt data structures. Standards enable systems providers and the manufacturing industry to speak the same language, facilitating integration and providing faster time to value.

For example, systems built on the international standard for integrating enterprise and control systems (ISA-95) normalize data across enterprise resource planning (ERP) and MES systems. Information at the MES level is very granular, focusing on factors that affect production output and quality (e.g., temperature and speed). However, ERP systems look at higher-level information, such as how long a given machine has been running. Manufacturing intelligence environments require the integration of both types of data.

It is also important to consider whether an environment has incorporated a generic data model that supports a hierarchical structure for reporting key performance indicators (KPIs) and metrics. Methods for establishing KPIs vary by industry. They should be set to help companies focus on what is urgent and what can be postponed by comparing a current state with a predetermined target.

Build in Flexibility with Open Systems: Manufacturing facilities are dynamic environments with new systems and equipment being added, modified, and removed over time. Systems in use differ from facility to facility, particularly in contract manufacturing arrangements. To readily accommodate such disparate and dynamic environments, manufacturing intelligence systems should be open and extensible.

Accelerate Payback: In today’s environment, life-science organizations cannot afford slow returns on their IT investments. When considering a manufacturing intelligence solution, companies should seek features that enable rapid roll-out and use. For example, prebuilt graphical dashboards and adapters for ERP/MES systems lower the total costs of ownership and shorten the time to real value.

Offer Something to Everyone: An IT system should provide benefits to multiple audiences — from the shop floor to the executive suite — with flexibility and configurability that enable different groups to get the information they need. A line operator, for example, should have access to raw information that allows him or her to detect problems in real time (e.g., a packaging line that is starting to overfill). A supervisor, however, should be able to access details that help reveal the cause of consistently lower than expected performance. Engineers leverage manufacturing data to determine how the age of equipment factors into its performance. Conversely, executives may look for higher-level trends to compare time, cost, and quality of manufacturing across lines, plants, or contract manufacturers. An enterprise manufacturing business intelligence environment should meet the needs of all these professionals and more.

Provide Actionable Information: All business intelligence solutions should deliver actionable intelligence. In a manufacturing intelligence environment, the software should enable users to clearly identify areas for improvement and yield insight into how to move forward.

The need for extended business intelligence in biomanufacturing environments has never been more acute. Manufacturers are squeezed by dwindling pipelines and reimbursements as well as expanded regulatory requirements. Many rely increasingly on contract manufacturers, which builds a new level of management complexity into their organizations. Although the need for expanded insight has never been greater, manufacturers must balance this requirement against the need to optimize existing IT investments. Time to value is also critical. With careful planning and adherence to best practices, pharmaceutical manufacturers can achieve all these goals — and ultimately drive new levels of quality, productivity, and performance in their manufacturing operations.

About the Author

Author Details
Arvindh Balakrishnan is vice president for the life sciences industry at Oracle Corporation, 500 Oracle Parkway, Redwood Shores, CA 94065; [email protected]. Karen Theel is senior director of manufacturing strategy for Oracle Corporation; [email protected].

REFERENCES

1.) Smith, A, and W. Swanton. 2003.Enterprise Manufacturing Intelligence (EMI) Projects Boast Six-Month Paybacks and Extensive Opportunities for Incremental Expansion and Benefits, AMR Research, Boston.