We're building an adaptive digital twin for bioprocesses — so biotech teams can scale, transfer, and operate with fewer surprises.
About
Bioprocesses are hard to run consistently. Conditions drift, equipment behaves differently across scales, and the gap between a small R&D batch and a production-grade run is where most of the firefighting lives.
Modica is developing an Adaptive Digital Twin for bioprocess modelling — a system that pairs with modular bioreactor hardware, dynamically reconfigures as the process changes, and aims to make scaling and near-continuous operation more predictable. It's early-stage R&D, funded in part by the EU through the FENG programme.
Vision
You need tools that reduce risk, save time, and fit your budget – not another dashboard that only experts can interpret. MODICA ADT turns bioreactor data into actionable guidance so your team can react before a batch goes off track or a fault blows the schedule.
Live data and early anomaly detection so you move from Quality by Design to Quality by Control (QbC) – fewer OOS, less stress at release and audit.
Our hybrid models work with limited historical data (Small Data), so you don't wait for huge datasets. Get predictions and recommendations earlier in development – and justify the investment with faster scale-up and fewer failed lab runs.
MODICA ADT is built for the QB Systems modular bioreactor ecosystem. The digital mirror updates within minutes after any config change, with no manual reconfiguration – so the twin always matches your lab setup and you're not gluing together separate systems.
Product
Less babysitting, fewer failed runs, and decisions your operators can act on – without a modelling expert in the loop.
Bioreactor–twin connection held for 48 h+ continuous operation (gaps <15 min) so your digital mirror stays online when the run is on.
≥95% sync of process parameters so reactor and model stay in lockstep – reliable basis for predictions and for release.
Hardware changes are detected automatically and the digital model updates in minutes – avoiding configuration errors.
Simulate process configurations and parameters without touching the reactor – fewer wet experiments, less material.
Real-time estimation of biomass and product concentration where you don't have physical sensors.
Test alternative parameters in dry runs – iterate faster and cheaper and reduce the number of failed lab trials.
Clear, actionable recommendations when the process drifts – so operators can respond with confidence.
Know what really drives performance and which parameters matter most for optimization.
Detect trends before they hit the batch; target: resolve in ~0.5 h (down from 30 h) to protect yield.
Three operational areas where bioprocess teams lose the most time, and where an adaptive digital twin can change the shape of the work.
We're an early-stage team — happy to talk to biotech and bioprocess operators about what we're building.
Get in touch