Pharma’s visual upgrade: the move to video-led investigation

25 5 minutes read

Listen to this article

Axis Communications’ Gabriele Mangiafico argues that pharmaceutical manufacturers should build visual oversight into their automation plans, giving earlier warning of disruption and clearer information when investigating deviations.

Recent research shows a pharmaceutical manufacturing sector under pressure. Rising operational costs are affecting 61% of those surveyed in pharma, above the average of the manufacturing sectors surveyed. Pharma respondents were also more likely than average to highlight productivity challenges, at 31%, while 39% identified regulation and compliance as a major operational test.

Automation and digitalisation offer one route to addressing these pressures. The controlled conditions, approved processes and extensive operational data of pharma production make it an ideal candidate for structured innovation. Yet only a quarter of those surveyed recognised the significant potential of network cameras to contribute visual data beyond security, which leaves a gap between the problems manufacturers want to solve and the technologies they recognise as useful.

Uniting the data of batch records, environmental monitoring, access logs and equipment outputs tells an incomplete story, as quality teams may still be unable to answer the basic question of what was physically happening in the event of a deviation. Greater automation and efficiency also cannot be allowed to reduce operational understanding.

Properly designed visual systems can help narrow this context gap, enabling pharma producers to connect analytics and visual information with existing production data while preserving the authority of validated process controls. Network cameras equipped with video analytics should therefore be considered a core part of the industry’s growing automation strategy.

Reconstructing the context of a deviation

The quality and compliance pressures identified in the research become most acute when a manufacturer has to explain a deviation. Pharmaceutical sites are inherently complex and tightly managed. Entry into controlled areas, the transfer of components and activity around production equipment may all be governed by standard operating procedures and backed up by written or electronic records.

Consider the investigation of an environmental monitoring alert within a controlled production area. Monitoring systems and sensors record the conditions around the incident and remain the authoritative source. Time-synchronised footage can then help the investigation establish the status of the area at the time, highlighting something like a door left open or unexpected movement nearby. These observations may support or eliminate investigative hypotheses, without replacing environmental or process data. Instead of reviewing an entire shift, the quality team can go directly to the relevant time and location.

Analytics can also draw attention to certain visible events before they are discovered during a later review, perhaps including activity within a restricted zone or movement across a defined boundary. They can generate an alert which carries the associated footage directly to an assessor. If the event has no bearing on the process, it can be closed. If it raises a concern, the appropriate procedure can begin while the details are still fresh. The result is a more focused investigation and a faster route from alert to assessment.

Adding visual information to validated systems

Visual information can reduce time spent on unrelated parts of an investigation, but to be clear, video plays a supporting role. It does not replace the strict systems and validated controls needed to release pharmaceutical products.

When footage is time-correlated with an alarm, access event or production timestamp, it can provide additional context for deviation management and become part of the documented information reviewed during an investigation. It may also help teams distinguish between an isolated occurrence and a process weakness that requires corrective or preventive action. This strengthens the evidence available when reviewing an event without changing the basis on which a product is released.

Visibility in a more automated plant

Pharmaceutical manufacturers were more likely than any other segment surveyed to cite the need to adopt automation and digitalisation, with 26% going as far as to identify it as a leading challenge. Wherever a business sits on that journey, greater automation changes how production is supervised. As routine intervention decreases and operators increasingly review by exception, maintaining clear oversight of equipment and processes becomes part of the automation decision itself.

Cameras can give engineers a live view before they enter controlled areas and help them examine the lead-up to a stoppage. Analytics can continuously monitor defined parts of the process and draw attention to events that meet set criteria, allowing operators to focus on exceptions rather than multiple live feeds.

Video also supports remote collaboration. Engineers, production specialists and quality teams can examine the same event without all entering the production area. Building visual oversight into the automation strategy can make fault-finding faster, reduce unnecessary intervention and ensure that reduced routine operator involvement does not leave the plant with a new blind spot.

Design for the production line

Like any other step in the pharmaceutical manufacturing process, the application of video should be carefully defined before it is deployed. There are physical limitations to consider: a camera intended to support an automated filling line will need a different view from one monitoring material transfer or access around an isolator, and aspects such as cleaning, mounting, cable routing and housing design bring the camera’s working environment into focus.

Logical aspects must be given equal weight. The event to be detected, the person who receives it and the expected response should be established before equipment is specified. The reality of the production floor does not always help. Protective clothing may affect how people appear to an analytic, while machinery can obscure important parts of the process. Lighting, reflections and restricted mounting positions may also limit the available view.

Cameras and analytics must be tested against the real scene and under realistic operating conditions. A specification cannot show how changing light, obstructed views or normal activity will affect detection. The expected event and the basis for judging performance must therefore be agreed before a trial begins.

Footage from pharmaceutical production areas may reveal proprietary processes as well as employee activity, so access, retention and cybersecurity policies must be established alongside the system. Its intended use should also be documented, particularly where visual information may inform an operational or quality decision. A camera must do its job – but not collect data beyond its defined operational purpose.

Cost, productivity and compliance already support the case for automation. The ability to supervise that automation should form part of the same business case. That decision belongs in the automation brief, alongside the systems that control and record the process. If visibility is left until after commissioning, the plant may inherit a blind spot that could have been designed out.

In Focus: A Spotlight on Manufacturing by Axis explores how visual intelligence can help manufacturers add operational context, close the visibility gap and support smarter, safer and more efficient manufacturing.

Gabriele Mangiafico, Business Development Manager, EMEA, Axis CommunicationsGabriele Mangiafico is part of Axis’ EMEA Business Development team, driving growth through digital transformation and innovative solutions in the industrial and manufacturing sectors. With a background that includes technical consultancy in NLP and computer vision, as well as business development roles for leading industrial companies, he has led complex projects and strategic roadmaps across the industry and is passionate about shaping the future of smart manufacturing.