RoslinCT Blog Series: Accelerating digitisation in cell & gene therapy manufacturing
Accelerating digitisation in cell & gene therapy manufacturing: Real-world insights and practical strategies from industry leaders
Cell and gene therapies have moved beyond the promise stage. Patients are already benefiting, indications are expanding, and more programmes are moving through development than ever before. But as the science proves itself, a new bottleneck has become impossible to ignore: industrialisation.
How do we take highly complex, personalised (and often time-critical) therapies and make them reliably, repeatably, and at scale without costs spiralling and timelines stretching?
That’s the backdrop for RoslinCT’s recent digitisation webinar, moderated by Rebecca Street-Howard (Head of Digital Operations, RoslinCT), with insights from Tour Ala (Roche), Chris Baldwin (Resolution Therapeutics), Dr Nicole Mather (IBM and Non-Executive Director at the Cell and Gene Therapy Catapult) and Peterjohn “PJ” Millan (Spark Therapeutics). What emerged was a set of grounded lessons about what’s actually moving the needle in real-world operations.
Here are the most practical takeaways.
Digitisation isn’t the goal – industrialisation is
It’s easy to talk about digitisation as a “transformation programme”. The panel repeatedly brought it back to something more direct: digitisation is a means to solve the industrialisation problem.
For advanced therapies, that problem has a few familiar faces:
- reducing cost without compromising quality
- shortening timelines to treatment (because patients can’t always wait)
- scaling capacity globally
- building robust, auditable processes that hold up in the real world
Digitisation matters because it helps teams move from heroic, manual execution to repeatable performance.
Collaboration is a competitive advantage right now
One of the most interesting themes was how openly the speakers challenged the idea that digital should be proprietary.
In cell and gene therapy, the supply chain is rarely confined to one organisation. Hospitals, logistics providers, manufacturers, quality teams, and technology vendors all touch the process. That means your “end-to-end” is only as strong as the handoffs between partners.
The panel’s shared view: in today’s CGT landscape, collaboration often creates more value than guarding every digital asset. Common standards, shared approaches, and pre-competitive alignment can reduce friction across the ecosystem, making it easier for everyone to deliver therapies reliably.
That doesn’t mean everything should be shared. The line becomes clearer when you ask: Is this truly a source of intellectual property, or simply foundational infrastructure? Algorithms, proprietary process knowledge, and novel product elements remain protected. But foundational interoperability (how systems exchange information and how identity is tracked) is where the industry benefits when it aligns.
“Brain-to-vein” thinking changes priorities
A subtle but powerful learning came from the discussion around timelines.
Many people talk about “vein-to-vein” time: from cell collection to infusion. But the panel expanded this into “brain-to-vein”, starting the clock earlier, from the moment a physician decides a patient may need therapy, through to actually securing the right clinical slot, then manufacturing, release, and delivery back.
Why does this matter? Because it shifts digitisation priorities away from “what looks advanced” and toward “what reduces delays and prevents avoidable disruption”.
When you view the whole patient journey as one connected flow, the most impactful digital interventions become clearer.
The biggest wins aren’t glamorous: scheduling, orchestration, exception handling
If you’re expecting the answer to be AI, the webinar offered a reality check.
The panel repeatedly highlighted “unsexy” systems as the ones delivering outsized impact today:
- digital scheduling and resource planning across clinical sites, manufacturing, QC and logistics
- real-time visibility for shipments and status (especially when timing is tight)
- exception management for when the process falls off the expected operating flow
- automation to reduce deviations and human error where appropriate
- interoperability that connects the ecosystem into a coherent workflow
The logic is simple: CGT is a make-to-order world. If one element slips, apheresis delay, a courier issue, or a manufacturing slot mismatch, the entire chain feels it. Digitisation helps teams anticipate issues earlier and recover faster, rather than reacting late when the cost of delay is highest.
Electronic Batch Records + a “digital backbone” can shift quality and speed together
One standout learning was the impact of building a connected digital backbone in GMP operations, including electronic batch records (EBRs) integrated with key supporting systems.
Why it matters: EBRs don’t just replace paper. When implemented well, they reduce errors, streamline review, and create the auditability the sector needs, while also accelerating batch release.
The panel shared results that will resonate with anyone trying to scale:
- substantial error reduction (by removing paper-related mistakes and improving control)
- major time savings for QA review effort
- the potential to accelerate approval timelines without changing the production schedule or staffing model
In a cost environment where labour is significant, and quality is non-negotiable, this is the type of “both/and” improvement the industry needs: better quality and faster release, together.
Chain of Identity is the “never event” problem, digitisation is the answer
For autologous therapies especially, Chain of Identity (COI) and traceability are existential. If a patient were ever to receive the wrong product, it could be life-threatening. And unlike some quality attributes, this isn’t something you can simply “test into” the product at the end.
The panel’s message was clear: digital traceability is essential.
Practical enablers included:
- barcode/RFID-style tracking through the chain
- authenticated handoffs and clear custody points
- end-to-end visibility)
- an operating model where COI/COC accountability is explicit, not assumed
A key insight here was that “end-to-end” expands over time. It starts within your own walls—but eventually must include partners, treatment centres, and logistics. That’s where shared standards and (likely) increasing regulatory direction will become critical.
AI is promising, but it relies on strong foundations
AI came up in the context of supporting teams with decision-making and orchestration, especially spotting when processes drift from the ideal workflow and helping humans respond faster.
But the speakers were pragmatic: AI won’t rescue poor data or disconnected systems. It amplifies what you’ve already structured. That means the path to useful AI starts with the fundamentals: clean data, mapped processes, integrated platforms, and agreed standards.
