Steve McCarthy, Vice President of Digital Innovation, Sparta Systems05.03.21
Proactive quality has been the ultimate goal of every quality organization in the life sciences industry for several years. The industry has defined the drivers an organization must meet to attain proactive quality as a differentiator. These include agile, scalable quality management systems (QMS) being patient-centric and bimodal—considering both current and future needs for meeting operating requirements.
But it’s important to remember proactive quality is the ultimate outcome, the destination. Advanced digital technologies such as artificial intelligence (AI), advanced analytics, and digital quality (often referred to as Quality 4.0) provide the vehicle that helps organizations reach that destination faster. Let’s look into Quality 4.0’s impact on achieving proactive quality, the applications of AI in the life sciences, and use cases for the AI-empowered quality system.
A Quality 4.0 Primer
We hear a lot about Quality 4.0 and its impact on transforming quality, but what is it?
A recent study by Boston Consulting Group (BCG), in partnership with ASQ and a Deutsche Gesellschaft für Qualität (DGQ) study, defined it simply as “the application of Industry 4.0 advanced digital technologies to enhance traditional best practices in quality.”
Quality 4.0 is among many developments giving rise to advanced manufacturing or the “factory of the future,” in which digitally enhanced manufacturing structures and processes increase productivity and flexibility in the factory and throughout the supply chain. Productivity is a top reason for implementing Quality 4.0, with BCG stating it has the potential to boost productivity by 10-15 percent. Alongside this productivity boost is the ability to react quickly to quality issues. The perceived importance to manufacturing reflects the visibility of quality value created on the shop floor.
Study participants point to predictive analytics, sensors, and electronic feedback loops as the most important technologies for driving impact. The study found more than 60 percent said predictive analytics will significantly affect quality performance and the bottom line within five years.
Clearing Up Common Misconceptions about Quality 4.0
There are some common misconceptions about Quality 4.0. A study from LNS Research addresses some of them, including:
AI is Prevalent in the Life Sciences: But What's Missing?
Deloitte released a paper in November 2020 in which it stated AI is most effective when deployed strategically across the entire enterprise, allowing a life science company to reap large benefits. For manufacturing, applying AI technologies can help reduce manual oversight in manufacturing operations and allow tighter control of quality and operating costs through proactive quality control. This is achieved by applying machine learning (ML) to assess manufacturing data across batches and lines as well as predict quality issues, automated batch release, and optimized yield and output.
We’re seeing a proliferation of AI and other innovative technologies across the life science industry. There is certainly a growing use of AI embedded in medical products. We’re also seeing it increasingly used in management of electronic health records.
But something's missing—the application of these powerful, digital technologies for the improvement of quality and compliance in the GMP arena and application in the quality management system itself.
How AI Is Revolutionizing the Quality Management System
Industry has made some progress toward proactive quality. However, one of the main reasons it hasn’t achieved this promise after years of defining and articulating it is many of these necessary technological advancements simply weren’t available in the GMP environment. This is only now being realized through AI and natural language processing (NLP) innovations and their application to the quality system discipline.
There are key business problems that bring opportunities for augmenting quality management through innovation. For example, there’s too much volume and not enough people. It’s often extremely difficult to wade through the huge volume of low-risk complaint or quality event records to effectively identify the few with potentially high-risk signals.
On top of this, triage and analysis of all this unstructured data—essentially text, sometimes in multiple languages—is difficult and subjective.
We as humans also have an unconscious bias. Sometimes the most experienced of us can suffer from that the most. We think we “just know the answer” because we’ve seen it so many times before.
The ability to leverage disruptive technology like AI unlocks the promise of proactive quality. AI in the QMS augments decision making, ultimately enabling predictive analytics with capabilities that enable firms to accomplish the following.
Quickly Classify and Triage Critical Quality Events: Automatic and systematic classifications point the AI engine at single records, individual complaints, or quality events. The algorithms then consume all the structured and unstructured data in that record, often large volumes of text entered by the quality professional. The AI then suggests the most likely categories of critical data elements within that record with an accompanying statistical probability—e.g., complaints type, reportability, and severity.
The remaining fields and requirements of the process and associated records then auto-populate on the users' acceptance of that suggestion. This augments human decision making to increase the timeliness of these critical actions, providing consistent categorization, improving signal detection, and ultimately the ability to prioritize your most valuable asset—people.
Automatically Correlate Quality Events: The AI engine uses NLP capabilities to rapidly identify patterns, trends, and occurrences of quality events across the entire event dataset. This includes complaints, nonconformances, deviations, corrective and preventive action (CAPA), investigations, change controls, and audit records.
It immediately shows the user-related and similar records at a click of a button, and clearly identifies the associated statistical relationship.
This capability helps suggest the “next best action,” resulting in faster investigations and superior trend analysis. It also significantly reduces redundancy and allows the ability to show regulatory authorities you have assessed the potential for a CAPA in one site, one product, and one process to be relevant to another.
Connect Operational and Quality Data: With Industrial Internet of Things (IIoT) technologies powering factories, manufacturing data can be connected to the QMS to detect anomalies in manufacturing operations to create and triage quality events to address potential issues in near real-time. This reduces the cost of poor quality by preventing distribution of products that may have resulted in a recall or patient harm.
The IIoT enabled capabilities also include a contextualized product disposition workflow to review and approve batches for market release. Users can achieve faster time to market by reducing the time required to review batches while maintaining the same quality standards.
The value and benefits of augmenting the expertise of quality teams with the speed and power of AI are clear. We can now perform consistent categorizations with timely reporting and risk assessment. We can carry out faster, more effective investigations and avoid redundancy.
We can also improve prioritization and view next best actions for quality events. Perhaps most importantly, we can optimize signal detection and trending, which minimizes errors, compliance risk, and associated regulatory actions.
Predictive analytics will reduce quality events over time. The associated productivity and ability to offset quality unit headcount increase with business growth is vital. Ultimately, this adds up to higher quality product, safer patients, a more secure supply chain, and unparalleled levels of global regulatory compliance.
Accelerate the Journey to Proactive Quality
Remember that proactive quality is the destination; Quality 4.0 is merely the vehicle. The advent of AI and IIoT capabilities within the quality system have brought us closer to reaching proactive quality.
The amazing power and value these technologies bring are clear. But this transformation won’t be easy. The key is culture and talent—the human element.
AI-empowered quality systems augment human decision-making to deliver the technology, maturity, and environment required to make such a paradigm shift and further improve the culture of quality.
With a BSc in Medical & Industrial Biology, Steve McCarthy is vice president of Digital Innovation at Sparta Systems. McCarthy is certified in six-sigma process excellence and provides domain expertise and serves as an industry evangelist and customer advocate for Sparta. McCarthy has nearly three decades of experience as a quality and supply chain leader within the healthcare industry.
But it’s important to remember proactive quality is the ultimate outcome, the destination. Advanced digital technologies such as artificial intelligence (AI), advanced analytics, and digital quality (often referred to as Quality 4.0) provide the vehicle that helps organizations reach that destination faster. Let’s look into Quality 4.0’s impact on achieving proactive quality, the applications of AI in the life sciences, and use cases for the AI-empowered quality system.
A Quality 4.0 Primer
We hear a lot about Quality 4.0 and its impact on transforming quality, but what is it?
A recent study by Boston Consulting Group (BCG), in partnership with ASQ and a Deutsche Gesellschaft für Qualität (DGQ) study, defined it simply as “the application of Industry 4.0 advanced digital technologies to enhance traditional best practices in quality.”
Quality 4.0 is among many developments giving rise to advanced manufacturing or the “factory of the future,” in which digitally enhanced manufacturing structures and processes increase productivity and flexibility in the factory and throughout the supply chain. Productivity is a top reason for implementing Quality 4.0, with BCG stating it has the potential to boost productivity by 10-15 percent. Alongside this productivity boost is the ability to react quickly to quality issues. The perceived importance to manufacturing reflects the visibility of quality value created on the shop floor.
Study participants point to predictive analytics, sensors, and electronic feedback loops as the most important technologies for driving impact. The study found more than 60 percent said predictive analytics will significantly affect quality performance and the bottom line within five years.
Clearing Up Common Misconceptions about Quality 4.0
There are some common misconceptions about Quality 4.0. A study from LNS Research addresses some of them, including:
- Quality 4.0 Is all about technology: In reality, Quality 4.0 is enabled by digital technology, but industrial transformation occurs when optimizing products, people, and processes.
- A Quality 4.0 strategy is separate from traditional quality: Quality 4.0 should be deployed as an integral part of a quality strategy and transform quality.
- Industry 4.0 is overhyped and there are no results: It certainly can be overhyped like all tech, but it is delivering real value for many manufacturers.
- Quality 4.0 Is for quality teams: The quality team must lead in tech to improve success and align to improve processes and people. For every person with a quality title, there are multiple others investigating Quality 4.0.
AI is Prevalent in the Life Sciences: But What's Missing?
Deloitte released a paper in November 2020 in which it stated AI is most effective when deployed strategically across the entire enterprise, allowing a life science company to reap large benefits. For manufacturing, applying AI technologies can help reduce manual oversight in manufacturing operations and allow tighter control of quality and operating costs through proactive quality control. This is achieved by applying machine learning (ML) to assess manufacturing data across batches and lines as well as predict quality issues, automated batch release, and optimized yield and output.
We’re seeing a proliferation of AI and other innovative technologies across the life science industry. There is certainly a growing use of AI embedded in medical products. We’re also seeing it increasingly used in management of electronic health records.
But something's missing—the application of these powerful, digital technologies for the improvement of quality and compliance in the GMP arena and application in the quality management system itself.
How AI Is Revolutionizing the Quality Management System
Industry has made some progress toward proactive quality. However, one of the main reasons it hasn’t achieved this promise after years of defining and articulating it is many of these necessary technological advancements simply weren’t available in the GMP environment. This is only now being realized through AI and natural language processing (NLP) innovations and their application to the quality system discipline.
There are key business problems that bring opportunities for augmenting quality management through innovation. For example, there’s too much volume and not enough people. It’s often extremely difficult to wade through the huge volume of low-risk complaint or quality event records to effectively identify the few with potentially high-risk signals.
On top of this, triage and analysis of all this unstructured data—essentially text, sometimes in multiple languages—is difficult and subjective.
We as humans also have an unconscious bias. Sometimes the most experienced of us can suffer from that the most. We think we “just know the answer” because we’ve seen it so many times before.
The ability to leverage disruptive technology like AI unlocks the promise of proactive quality. AI in the QMS augments decision making, ultimately enabling predictive analytics with capabilities that enable firms to accomplish the following.
Quickly Classify and Triage Critical Quality Events: Automatic and systematic classifications point the AI engine at single records, individual complaints, or quality events. The algorithms then consume all the structured and unstructured data in that record, often large volumes of text entered by the quality professional. The AI then suggests the most likely categories of critical data elements within that record with an accompanying statistical probability—e.g., complaints type, reportability, and severity.
The remaining fields and requirements of the process and associated records then auto-populate on the users' acceptance of that suggestion. This augments human decision making to increase the timeliness of these critical actions, providing consistent categorization, improving signal detection, and ultimately the ability to prioritize your most valuable asset—people.
Automatically Correlate Quality Events: The AI engine uses NLP capabilities to rapidly identify patterns, trends, and occurrences of quality events across the entire event dataset. This includes complaints, nonconformances, deviations, corrective and preventive action (CAPA), investigations, change controls, and audit records.
It immediately shows the user-related and similar records at a click of a button, and clearly identifies the associated statistical relationship.
This capability helps suggest the “next best action,” resulting in faster investigations and superior trend analysis. It also significantly reduces redundancy and allows the ability to show regulatory authorities you have assessed the potential for a CAPA in one site, one product, and one process to be relevant to another.
Connect Operational and Quality Data: With Industrial Internet of Things (IIoT) technologies powering factories, manufacturing data can be connected to the QMS to detect anomalies in manufacturing operations to create and triage quality events to address potential issues in near real-time. This reduces the cost of poor quality by preventing distribution of products that may have resulted in a recall or patient harm.
The IIoT enabled capabilities also include a contextualized product disposition workflow to review and approve batches for market release. Users can achieve faster time to market by reducing the time required to review batches while maintaining the same quality standards.
The value and benefits of augmenting the expertise of quality teams with the speed and power of AI are clear. We can now perform consistent categorizations with timely reporting and risk assessment. We can carry out faster, more effective investigations and avoid redundancy.
We can also improve prioritization and view next best actions for quality events. Perhaps most importantly, we can optimize signal detection and trending, which minimizes errors, compliance risk, and associated regulatory actions.
Predictive analytics will reduce quality events over time. The associated productivity and ability to offset quality unit headcount increase with business growth is vital. Ultimately, this adds up to higher quality product, safer patients, a more secure supply chain, and unparalleled levels of global regulatory compliance.
Accelerate the Journey to Proactive Quality
Remember that proactive quality is the destination; Quality 4.0 is merely the vehicle. The advent of AI and IIoT capabilities within the quality system have brought us closer to reaching proactive quality.
The amazing power and value these technologies bring are clear. But this transformation won’t be easy. The key is culture and talent—the human element.
AI-empowered quality systems augment human decision-making to deliver the technology, maturity, and environment required to make such a paradigm shift and further improve the culture of quality.
With a BSc in Medical & Industrial Biology, Steve McCarthy is vice president of Digital Innovation at Sparta Systems. McCarthy is certified in six-sigma process excellence and provides domain expertise and serves as an industry evangelist and customer advocate for Sparta. McCarthy has nearly three decades of experience as a quality and supply chain leader within the healthcare industry.
