Issue link: https://www.e-digitaleditions.com/i/1495138
12 BioPharm International ® Quality and Regulatory Sourcebook March eBook 2023 www.biopharminternational.com compliance audit trail can include records of all equip- ment, software, materials, and reagents used in the laboratory, as well as records of equipment mainte- nance and calibration. Other important components include records of all data and results obtained, as well as any deviations from standard procedures and any corrective actions taken. Impact in the QC laboratory Intel l igent i n for mat ics ca n be used to automate many of the tedious and error-prone tasks involved i n d r ug development a nd produc t ion. Improved usabilit y of AI-powered sof t ware can help reduce the risk of errors and can make testing easier and more accessible than before. This improvement in accessibility can help reduce the likelihood of errors occurring during these processes, as well as improve efficiency and speed. The long-term value of intel- ligent informatics lies in the ability of scientists to use their time effectively to collect data by removing time-consuming manual tasks and allowing them to make better and faster business decisions (4). The goal is not instr umentation that thinks for itself, but technolog y that is intelligent enough to help humans do their jobs better and faster. This capa- bilit y generates productivit y gains and improves efficiency in the lab, while also allowing scientists to spend more time on value-added tasks such as method development and validation, data analysis, and data interpretation. System errors and checks. Smart technology in combination with instrumentation could f lag po- tential issues or errors, which may result in a failed sample ana lysis and lost productivit y, before the analyst starts an analytical run. The goal is to pre- vent common errors such as insufficient solvent to complete the run, use of the incorrect mobile phase or column, or selection of the incorrect method for a particular sample. This capability is important as laboratories hire more employees without extensive experience operating specific instrumentation or troubleshooting issues. Algorithms can also support checks for infrequent or uncommon errors that re- quire more advanced knowledge. Additionally, all data would be logged to learn about errors and even predict future ones before they occur. Lab management efficiency. AI-powered intel- ligent informatics offer opportunities to minimize wasted time for troubleshooting, training, data anal- ysis, and other activities that can take time out of the working day. Algorithms can analyze large amounts of data and identif y patterns that might not be im- mediately apparent to humans, helping to identif y potent ia l sa fet y issues before a produc t reaches the market. Because laboratories generate massive amounts of data, just finding those errors can take a lot of time, sometimes requiring days or weeks before the root cause is identified. That could have signif- icant implications if a product was released during that time, potentially causing the need for a recall or other costly remediations. Real-time error information. Many human errors are simply mistakes that technicians don't neces- sarily know they are making at the time. Currently, QC tec h n ic i a n s re ly heav i ly on over sig ht f rom other colleagues physically present in the lab—an aspect that has been particularly challenging with the increase in hybrid and remote work during the pandemic. Coupled with high turnover rates and onboarding challenges in the current employment market, many pharmaceutica l compan ies would benefit from AI algorithms that could report errors as t hey a re made, giv ing t he a na lyst a cha nce to learn from their mistake, as well as take steps to fix the error immediately. Training. I ntel l igent i n for m at ics ca n prov ide user training to tutor new technicians as they are completing tasks. Training may include visual in- structions for completing tasks, as well as guided workf lows that would traditionally be the responsi- bility of a lab manager or supervisor. This capability could work in combination with health checks and real-time error information to get new users up to speed as quickly as possible. Audit trails. Intelligent informatics can provide support for the necessar y audit trails that include i n st a l lat ion , h a rdwa re mod i f icat ion s, sof t wa re modifications, and instrument events. Pharmaceu- tical companies face substantial financial impacts if they are unable to quickly provide regulators with the required data, including documented evidence of compliance. Outdated technolog y and insuf f i- cient support can put data and compliance at risk, resulting in disruption, unplanned downtime, and other consequences. Conclusion While the benef its are numerous, adoption of in- tel l igent i n novat ions i n a laborator y set t i ng ca n be challenging due to the lack of familiarit y with Qualit y Control Effective change management requires the active participation of all stakeholders.