Connecting a hearing research CRO across three continents

What we'd propose

• Digital CDMO

  Real-time monitoring and digital twin for the OrgaEar bioprinting line

  An Industry 4.0 layer over the OrgaEar 3D bioprinting line: OPC UA acquisition from bioprinters, computer-vision quality monitoring during extrusion, and a digital twin (a live simulation of the process used to test changes before they are run on real tissue) of the organoid production process used to simulate parameter changes before they are run on real tissue.

  • Bioprinter data acquisition

    Reading every run as it happens

    Bring process values, actuator states and environmental conditions from each 3D bioprinter into a time-series layer using OPC UA, so a deviation is recorded against the specific run that produced it rather than inferred from a downstream assay.

  • Computer vision QC over the extrusion process

    Non-contact inspection during extrusion

    Apply camera-based OpenCV (an open-source computer vision library) processing to detect anomalies in the printed structure during extrusion, without physical contact with the medium, so a failed print is caught before the next layer is laid down.

  • Organoid process digital twin

    Simulate before running real tissue

    Build a digital twin of the inner ear organoid production process so a candidate parameter change can be simulated for its effect on organoid maturation before it is run on a real print, reducing wasted tissue and accelerating R&D cycles.

  • Each run is captured as a digital record rather than reconstructed from notes after the fact.
  • Failed prints are caught during extrusion rather than at the maturation assay, cutting material waste.
  • Parameter changes are simulated in the digital twin before they are committed in tissue, shortening development cycles.

  We've done this before

  • Bioreactor Control: Computer Vision for Non-Invasive Foam Management
  • An Adaptive Digital Twin for Bioprocess Autonomy

  See how we do Digital CDMO
• Enterprise AI

  Sensor-to-scientist data pipeline for auditory analytics

  An automated data pipeline that streams electrophysiology and clinical data from laboratory instruments into the AI training environment used by CILcare's Auditory Analytics department, replacing manual file transfers with validated ingestion at the source.

  • Instrument connectors and drivers

    Bringing instruments online

    Build custom drivers and connectors for ABR and DPOAE equipment and for clinical audiometry systems so data is captured with instrument identity, method version and timestamp attached rather than exported manually from each instrument's workstation.

  • Unified data lake for sensory and clinical data

    One source of truth

    Ingest sensory data, offline analytical results and clinical audiometry records into a shared data layer using a time-series store, so the Auditory Analytics team can query one dataset instead of reconciling extracts across workstations.

  • AI-ready data governance

    Datasets the model can trust

    Implement automated validation, normalisation and versioning of the datasets that feed AI models, so a model trained on data from Montpellier can be re-validated on data from Boston and Japan against an unambiguous lineage.

  • The Auditory Analytics team works from a live dataset rather than from manually assembled extracts.
  • Model validation across Montpellier, Boston and Japan reuses the same lineage rather than being repeated per site.
  • Data shared with Shionogi for milestone validation carries its own provenance back to the instrument that produced it.

  We've done this before

  • Ontology-Driven Ecosystem for Biomanufacturing Data
  • Transforming Raw Bioprocess Data into Actionable Insights

  See how we do Enterprise AI
• Digital Lab

  Greenfield digital laboratory architecture for the Japan subsidiary

  A reference architecture and equipment specification package for the Japanese subsidiary scheduled for 2026: MTP-based modular equipment onboarding, documented IT/OT reference architecture and segmentation rules so the new site's data path is designed rather than retrofitted.

  • MTP-based modular lab architecture

    Plug-and-produce equipment onboarding

    Specify equipment and integration requirements around MTP so instruments can be added or replaced without a custom integration project each time, and the Japan site opens with the same equipment onboarding shape as Montpellier and Boston.

  • IT/OT reference architecture for the site

    One documented data path

    Specify how laboratory equipment, line supervision and enterprise systems connect at the Japan site, including the segmentation model and protocol choices, so every vendor on the project builds toward the same target.

  • Transatlantic and trans-Pacific replication backbone

    Data moves on a known path

    Define the secure replication backbone between Montpellier, Boston, Copenhagen and Tokyo, with conflict resolution and regulatory-compliant partitioning, so partner data exchange with Shionogi happens over a known channel rather than ad hoc.

  • The Japan subsidiary opens with a data path designed for Shionogi milestone reporting from day one.
  • Equipment onboarding follows MTP, so adding an instrument is a configuration step rather than an integration project.
  • The same architecture and procurement language is reusable for future CILcare sites in the same network.

  We've done this before

  • Future-Proofing R&D: Migration to High-Availability Cluster Architecture
  • Restoration: Establishing or Re-establishing a Complex Lab Ecosystem

  See how we do Digital Lab
• Digital Lab

  ALCOA+ audit trail across preclinical and clinical studies

  A Laboratory Execution System integrated with the existing LIMS and ELN that generates ALCOA+ audit trails as a by-product of laboratory work, taking the manual reconstruction step out of FDA and EMA submissions for the auditory signature biomarker recognition programme.

  • Laboratory execution system

    Orchestrating people, instruments and IT

    Implement an LES that orchestrates people, instruments and IT systems into a single workflow, replacing paper-based toxicology and study reports with automated digital records that capture every step in order.

  • ALCOA+ audit trail generation

    Evidence produced as the work happens

    Generate the audit trail automatically as experiments are recorded, ensuring every data point is Attributable, Legible, Contemporaneous, Original and Accurate, with the additional ALCOA+ requirements of Complete, Consistent, Enduring and Available, so the evidence for a regulatory submission is produced by the system rather than compiled afterwards.

  • Biomarker submission data packaging

    Packaging data for FDA and EMA

    Build automated data packaging workflows that prepare auditory signature datasets for submission to the FDA and EMA, ensuring the digital biomarker evidence meets the regulator's expectations for clinical validation.

  • FDA and EMA inspections are answered from the system record rather than from a reconstruction.
  • Submission packaging for digital biomarker recognition becomes a data export rather than a documentation project.
  • GLP environments move from hybrid paper and digital records to a fully inspectable electronic chain.

  We've done this before

  • Digital Transformation of QC Labs: From Paper-Based Legacy to Integrated LES
  • Creating a Digital Lab for a Global Biotech Player

  See how we do Digital Lab
• Enterprise AI

  IT/OT segmentation for secure partner data exchange

  A network and identity architecture for CILcare's transatlantic and Japan-bound data flows, designed so proprietary AI models and drug-candidate data can be shared with Shionogi and with CILcare's subsidiaries on a channel that does not expose the rest of the network.

  • IEC 62443 zones and conduits

    Segmentation designed before traffic scales

    Define zones, conduits and remote-access rules to IEC 62443 across the Montpellier, Boston, Copenhagen and Japan sites, so laboratory OT (Operational Technology — the sensors, controllers and instruments that run the lab) is isolated from corporate IT (Information Technology — the office and enterprise systems) and from partner traffic.

  • Zero Trust access for R&D data

    Identity-based access to models and records

    Implement identity-based access control for proprietary AI model repositories and drug-candidate records, so access is granted per user and per resource rather than per network location, and partner access can be scoped to specific data sets.

  • Secure partner data exchange channel

    Shionogi collaboration on a known channel

    Deploy an encrypted data exchange channel between CILcare and Shionogi that meets Japanese and European data sovereignty requirements, so collaboration on the CIL001 and CIL003 milestones happens over a documented channel that satisfies both regulatory frameworks.

  • Partner collaboration with Shionogi happens on a channel that satisfies both Japanese and European data sovereignty requirements.
  • Proprietary AI models and drug-candidate data are isolated from broader network exposure.
  • The segmentation model is designed once and reused as the network grows to Japan and beyond.

  We've done this before

  • Scaling Biotech Manufacturing via Universal IT/OT Architecture

  See how we do Enterprise AI