COVID-19 Chronicles: Bring Automation to Continuous Manufacturing in the Pharmaceutical Industry with Data Analytics-as-a-Service

The year 2020 has become a year of economic instability, industrial restructuring, consumer uncertainty, and global pandemic. COVID-19 is impacting various industries, showcasing supply chains’ lack of resilience and agility and challenging the global overproduction and consumerism culture. Despite pressure and uncertainty, the pharmaceutical, healthcare, and technology industries are the most critical industries of the COVID-19 world. Having experienced substantial growth during COVID-19, the pharmaceutical industry is undergoing a transition toward automation and continuous manufacturing. The pandemic is accelerating standards, data-enabled decision making, and Machine Learning (ML)-driven drug discovery.

The pharmaceutical industry and the chemical production, biopharma, and bio-med domains are strongly interconnected with the Internet of Things (IoT), data-enabled technologies, and data-driven applications. Since the rapid digitalization and automation of the pharmaceutical industry, the volume and variety of data have grown dramatically, especially in the medical-application domain. Nevertheless, the pharmaceutical industry is at the very beginning of its move toward continuous manufacturing and scaling up its data monitoring systems. The U.S. Food and Drug Administration (FDA) only officially encouraged and published the guidance for the adoption of continuous manufacturing for pharmaceutical enterprises in 2019. Taking pressure from COVID-19 into consideration, there is a critical need to implement continuous manufacturing processes, specifically to reduce manufacturing time, minimize the risk of human error, and further enable advanced monitoring systems (monitoring techniques, anomaly detection, and predictive maintenance).

As a result of implementing continuous manufacturing in the pharmaceutical and chemical industries, there will be a need for a better and more sophisticated control system. These systems would be required to efficiently capture, process, and integrate data coming from the production floor and various mission-critical pieces of machinery. Learning from the continuous manufacturing in the industrial domain, the pharma industry would be challenged with integrating data across different assets as well as monitoring Key Performance Indicators (KPIs) for continuous pharmaceutical processes in near real time. Then, process automation with advanced analytics of the obtained data, application across various functions and solutions, ML, and Artificial Intelligence (AI) would be further ideas to consider. Finally, COVID-19 has highlighted weaknesses in terms of agility and resilience for pharma and other industries. Hence, fast-tracked digitalization and analytics automation tooling would be key to agility acceleration and enabling transparency and the optimization of the processes, which, in retrospect, could be enhanced by the transition toward continuous manufacturing

Due to the novelty and high degree of industry regulations, vendors with analytics and data-enabled technologies are scarce. Currently, Seeq is making waves in the pharma industry with its specialized solution and narrowed down pharma-related applications. Seeq is a software company that was founded in 2013 and is headquartered in Seattle, Washington in the United States. The company provides software applications for advanced analytics and data computation, enabling interactive, visual tools to accelerate industrial process analytics, with a focus on continuous manufacturing with significant clients in the pharma industry (such as Merck, Vertex, Sanofi, Abbot Nutrition, and Lonza). The most popular applications provided by Seeq in pharma, chemicals, and life science verticals are batch vessel monitoring, continued process verification for continuous manufacturing, streaming, and advanced analytics for continuous biomanufacturing.

Seeq’s software kit includes three main components: Organizer, Workbench, and Data Lab:

• Seeq Workbench is automated data preparation, cleansing, and contextualization of the data analytics component of the software. The Workbench includes capabilities such as profile search, low pass filter, and scaled calculation. Seeq is offering those operational AI and ML capabilities under the hood via Workbench capabilities. The software enables advance compute to be performed with abilities to deploy predictive maintenance and diagnostic modeling as well as increase data transformation and readability. While Seeq’s Workbench is not positioned in the AI/ML market directly, its capabilities and functionality are resonating with the capabilities of operational AI and ML.

• Seeq Data Lab is Seeq’s newest application or data scientists and process engine, which is based on proprietary Python Libraries (SPY) and builds on Jupiter Notebooks. This product is a “typical” cloud-centric ML/AI with more traditional load-store-analyze data-at-rest options. The application is directed toward data scientists and Information Technology (IT) professionals to execute custom python programming by deploying any algorithms. The most used Data Lab solutions are the diagnostics, which enable the exposure of causality and influence the data discovery, as well as a monitoring function, which finds hidden trends and detects the anomalies and predictive compete with abilities to predict actions before the impact.

However, after taking a precise look at the industry challenges, it is fair to suggest that continuous manufacturing is the goal, but not the means. What the pharma industry is primarily craving is the consolidation and integrity of the data, which can then be utilized to improve the efficiency of continuous manufacturing. The FDA took a step further and even described and regulated the data integrity challenge for the pharmaceutical industry, where it is defined as “an assurance that data records are accurate, complete, consistent, and maintained within their original context, including their relationship to other records.”

Since the acceleration of remote-everything, the challenge of data integrity is going further than directly transferring records in the electronic system. The technology and the IoT Data Analytics-as-a-Service (DAaaS) are stepping up to eliminate the human error from all the points across the data value chain, such as data collection, ingestion, processing, and storage. The transition to viewing data from a holistic approach, where the data is generated from various resources and manufacturing functions, nevertheless still results in data silos and fragmentation of the analytical resources. Therefore, the data integrity challenge and need for automation for continuous manufacturing allows companies like Seeq to make moves and disrupt the industry with their technology. Since Seeq can “connect” to any data source the company is presented with, the data consolidation challenge can be quickly addressed. Alongside the ability to continuously update, analyze, and centralize the data, Seeq also enables advanced analytics outputs to automate continuous manufacturing decision making and monitoring of assets.

All in all, the pharmaceutical industry is at the beginning of ensuring total data consolidation and adopting continuous manufacturing, with future plans of automating decision making. However, ABI Research expects that IoT analytics as a service in combination with the pandemic and established base of the regulation regarding data would accelerate the adoption of the technology.