Temporal Analysis of Biomanufacturing Skill Development: Understanding the Needs of an Industry-4.0 Workforce

HTTPS://STOCK.ADOBE.COM

The transition toward industry 4.0 represents a significant evolution in biomanufacturing, driven by the integration of advanced digital technologies in four technological domains: data analytics, automation, cybersecurity, and intelligent sensor systems. Herein, we identify skills gaps in those domains, all of which will be critical to harnessing the full potential of biomanufacturing 4.0. The data analyzed represent demand frequencies over time for technical skills that are necessary for optimizing biopharmaceutical production processes. Understanding such trends in skills supply and demand will be essential in developing targeted educational programs and industry initiatives that can address future workforce needs.

For each technological domain, we analyze the frequency of technical skills in biomanufacturing-sector job postings over five years. Comparing skills that were in demand during 2017 with those at the beginning of 2023 can yield valuable insights into the evolving landscape of biopharmaceutical operations. Key trends include the rise of machine learning (ML) and cloud-based solutions in the domain of data science, integration of automation technologies into the internet of things (IoT), application of artificial intelligence (AI) in cybersecurity, and development of advanced sensor technologies.

Our Method

To analyze job-skills and domain-knowledge requirements, we collected data from 2017 to 2023 using two functions in the Lightcast labor-market database. The Job Posting Analytics (JPA) function covers about 165.6 million online job postings from >130 different online sources (e.g., LinkedIn, Indeed) in the United States. JPA data cover skills requirements in the labor market. The Profile Data (PD) function provides public, self-reported information about individuals’ job histories, educational histories, and expertise.

The Lightcast database retrieves real-time labor-market data and applies built-in data-mining methods to identify labor-market patterns and frequency of occurrences for specific skill-describing keywords/phrases. The JPA and PD functions support customized searches through different filters. For instance, results can be filtered by keywords in full job postings. For our analysis, job-posting and profile data were filtered, ranked, and sorted by skills, job title, region, time frame, and source industry.

Note that the number of postings listed in the Lightcast database does not equal the number of hires, nor does it equal the number of positions that are actually available. Postings for a given set of jobs can outnumber hires when a company is trying hard to find talent. Posting numbers can be significantly lower than the number of hires in certain job types that typically are not advertised online, as is the case with some “blue-collar” positions. In addition, the same job posting can be listed on multiple websites. The JPA function removes duplicates across sources.

The Lightcast PD database, which contains information from about

143.4 million US profiles, covers 86% of the American workforce. Considering the nature of the Lightcast database, we presume that information from the PD function realistically represents the distributions of skills and domain-knowledge topics in job postings and online profiles.

We categorized, classified, and organized profile and job-posting data from 2017 to 2023 by skills, job titles, regions, time frames, and source industry. Where possible, sample data were supplemented and cross-checked using databases from LinkedIn and numerous universities.

Analytical Framework: After defining our four technology domains, we extracted data about skills and qualifications (labor supply). For each domain, we applied specific keywords in the Lightcast database JPA module to generate lists of related skills and qualifications. Here are some examples:

• data science — data mining, big data, programming languages, AI, machine learning, mathematical and statistical skills (1)

• automation — automation, robotics, lean manufacturing, simulation, root-cause analysis, quality management, continuous improvement (2)

• cyber capabilities — cybersecurity, agile, cloud computing, IoT, web development, solution architectures, development opportunities (3)

• sensors — sensor analytics, signal processing, process analytical technology, digital twin, spectroscopy, optical, electrochemical, continuous manufacturing, batch manufacturing (1).

Next, we analyzed job descriptions, extracting and evaluating titles, skills, and qualifications requested by biomanufacturing companies. Titles that appeared in the database <1000 times during the study period were excluded. Then, we compared job-postings and job-profile data to identify gaps between required (demand) and available (supply) skill sets. We also tracked demand frequency for the top five skills in each technology domain over the five-year analysis period. Subsequently, we developed linear-regression models to project skills demand in biomanufacturing from 2024 to 2030 (data not shown).

As we mentioned, identifying demand for specific skills at the beginning and end of our analysis period (2017 to 2023) can yield valuable insights into the evolving landscape of biopharmaceutical operations. For instance, temporal analysis of job-skill demand highlights technological advancements and priority shifts that have occurred within the industry, reflecting companies’ responses to new challenges and opportunities. 

Key Insights

Data Science: In 2017, the terms listed most frequently in biomanufacturing-industry job posts for data-science positions were Python, SQL, R, big data, and data mining (Figure 1). Experience with the Python programming language ranked highest in demand frequency, receiving mention in a little under 35% of data-science job postings. Coming into 2023, the Python, SQL, and R languages remained critical areas of interest, with demand frequency continuing to increase. Such results indicate those languages’ foundational role in data manipulation and analysis. Python has become especially dominant because of its versatility and extensive libraries for data science, ML, and automation.

Figure 1: Frequency in job postings for five highly demanded data-science skills in 2017, measured again in 2023 (raw data from the Lightcast database).

The increasing prominence of ML in job postings — e.g., for experience with the TensorFlow and PyTorch ML platforms — signifies growing interest in advanced analytics techniques and predictive modeling. Furthermore, data-visualization tools such as the Tableau platform are becoming critical as biopharmaceutical companies demand deep insights from sophisticated data models. Such trends highlight the need for data scientists to be adept in both traditional and modern data tools.

Demand is growing similarly for experience with “big data” and data warehousing, driven by the need to manage and analyze vast amounts of bioprocess data. Thus, the biopharmaceutical industry is undergoing a noticeable shift to cloud-based data systems. Amazon Web Services (AWS) has emerged as a key platform, as is evident from an increase in demand for AWS expertise: from 0% in 2017 to 70% in 2023. Companies want to adopt scalable and flexible data-management systems.

Automation: Although traditional skills related to programmable logic controllers (PLCs) and robotics continue to be important, demand is surging for experience with more advanced, interconnected technologies (Figure 2). Supervisory control and data acquisition (SCADA) systems and the IoT have gained prominence, confirming the industry’s movement toward “smart” manufacturing environments. Along similar lines, experience with Python has emerged as a vital skill, with demand climbing from 35% to 70% over the measurement period. That increase relates to the language’s versatility for automation scripting, process control, and process/system maintenance. Such evolutions point to a future in which automated biomanufacturing technologies are dynamic, data-driven, and integrated with broader IT infrastructures.

Figure 2: Frequency in job postings for five highly demanded automation-related skills in 2017, measured again in 2023 (raw data from the Lightcast database); IoT = internet of things, PLC = programmable logic controller, SCADA = supervisory control and data acquisition.

Cybersecurity has changed substantially even over the past five years, with a pronounced shift from network and endpoint security to sophisticated, holistic approaches. Traditional network and endpoint mechanisms remain critical, but there is an increased emphasis on cloud security and AI-based security, demand for which increased from 35% to 60% and from 0% to 70%, respectively (Figure 3). Those trends reflect the increasing complexity and sophistication of cyber threats in the biopharmaceutical sector.

Figure 3: Frequency in job postings for five highly demanded cybersecurity skills in 2017, measured again in 2023 (raw data from the Lightcast database); AI = artificial intelligence.

Biopharmaceutical companies’ introduction of blockchain security systems indicates industry awareness that new technologies bring potential vulnerabilities. The emergence of blockchain and the continued importance of security information and event management (SIEM) methodologies demonstrate that the industry is being proactive about securing sensitive data and systems. Cybersecurity professionals at life-science companies will need to update their skills continually to stay ahead of emerging threats. Furthermore, skills in penetration testing, cybersecurity frameworks such as NIST/ISO 27001 (4, 5), and threat intelligence are becoming more critical to mitigating the evolving threat landscape effectively.

In the sensors domain, demand for expertise mirrors recent advancements in sensor capabilities and integration needs. Traditional skills such as facility with sensor calibration and embedded systems remain important. But biopharmaceutical companies increasingly are emphasizing IoT integration (from 30% to 60%) and wireless communication (from 35% to 65%) (Figure 4). Rising demand for experience with ML and sensors (from 25% to 55%) further indicates the industry’s interest in using smart sensor systems with (at least some) autonomous capabilities. Additionally, edge computing, sensor fusion, and advanced signal-processing techniques are becoming increasingly important to support the growing complexity and functionality of modern sensor systems.

Figure 4: Frequency in job postings for five highly demanded sensor-related skills in 2017, measured again in 2023 (raw data from the Lightcast database); IoT = internet of things.

Additional Observations

Demand for AI and ML skills has grown rapidly. That trend underscores the biopharmaceutical industry’s push to leverage such capabilities for data analysis, predictive modeling, and automation. Cloud technologies such as the AWS platform also have received increasing attention, reflecting companies’ shift toward scalable storage solutions, increased computational power, and improved data management.

A rise in demand for experience with software-development platforms such as those offered by Docker, Jenkins, and GitHub indicates biopharmaceutical-industry adoption of software development–IT operations (DevOps) practices to streamline development processes, improve collaboration, and enhance continuous integration and delivery pipelines. The Apache Kafka, Apache Spark, and NoSQL databases are gaining industry traction, highlighting a need to handle large volumes of data efficiently and to derive actionable insights from complex data sets.

Finally, while demand for traditional coding languages such as C and C++ has declined, Python is dominating because of its versatility and extensive libraries for data science, ML, and automation.

Planning for an Industry-4.0 Workforce

Our analysis reflects the dynamic nature of the biomanufacturing industry, in which rapid technological advancements and changing priorities drive the evolution of skills across scientific domains. The data presented herein illustrate the evolution of key technical skills across four technological domains associated with the biomanufacturing 4.0 paradigm. Comparing the frequency of skills in those domains from 2017 with values leading into 2023 highlights significant shifts in the industry’s skill demands. Our analysis aims to shed light on technological advancements and changes in hiring priorities for biopharmaceutical operations.

The increasing importance of AI, ML, the IoT, cloud computing, and other such advanced technologies is evident across all four domains. Such changes underscore the industry’s commitment to innovation and efficiency, driven by the need to manage complex data, enhance automation, ensure robust cybersecurity, and integrate sophisticated systems for process monitoring. As biomanufacturing continues to evolve, demand for those skills is likely to grow, necessitating ongoing investment in workforce development and training that emphasize those capabilities.

References

1 Li G, et al. Data Science Skills and Domain Knowledge Requirements in the Manufacturing Industry: A Gap Analysis. J. Manufact. Sys. 60, 2021: 692–706; https://doi.org/10.1016/j.jmsy.2021.07.007.

2 Erickson J, et al. End-to-End Collaboration To Transform Biopharmaceutical Development and Manufacturing. Biotechnol. Bioeng. 118(9) 2021: 3302–3312; https://doi.org/10.1002/bit.27688.

3 Pandey M, Singhal B. Blockchain Technology in Biomanufacturing: Current Perspective and Future Challenges. Blockchain Technology for Emerging Applications: A Comprehensive Approach. Hafizul Islam SK, et al., Eds. Academic Press: Cambridge, MA, 2022: 207–237.

4 NIST Cybersecurity Framework 2.0. US National Institute of Standards and Technology: Gaithersburg, MD, 26 February 2024; https://nvlpubs.nist.gov/nistpubs/CSWP/NIST.CSWP.29.pdf.

5 ISO/IEC 27001:2022 (3rd edition). Information Security, Cybersecurity, and Privacy Protection — Information Security Management Systems — Requirements. International Organization for Standardization: Geneva, Switzerland, 2022; https://www.iso.org/standard/27001.

Corresponding author Jason Beckwith, PhD, is managing director of Evolution Search Partners Ltd. in Glasgow, UK, and a researcher in the department of business at the University of Dundee; 44-7899 950002; [email protected]. William Nixon, PhD, is a emeritus professor, and Stavros Kourtzidis, PhD, is deputy associate dean of research, both in the department of business at the University of Dundee. Stephen Goldrick, PhD, is an associate professor in the department of bioprocess engineering at University College London. The authors report no declarations of interest.