☰

European Biomedical Institute (EBI)

Scaling biocompatibility testing for ISO 2025

Industry
Contract Research Organization (Medical Device Biocompatibility Testing)
Headquarters
Józefów, Poland
Public information as of
January 2026

A4BEE prepared this analysis from publicly available sources. It reflects our own reading of European Biomedical Institute (EBI)'s published strategy and is not endorsed by, or produced in cooperation with, European Biomedical Institute (EBI). Company website

Strategic priorities

The European Biomedical Institute, legally Europejski Instytut Biomedyczny Sp. z o.o., operates as a contract research organization specialised in medical device biocompatibility testing under ISO 10993. Its primary laboratory sits in Józefów, south of Warsaw, with supporting operations through NABI in the United States and Konmex Limited in the United Kingdom. The institute's stated positioning is the shortest possible lead time in biocompatibility evaluation at the highest quality of testing.

The October 2025 revision of ISO 10993-1:2025 sets the present planning horizon. The new standard expands Biological Evaluation Plan (BEP) and Biological Evaluation Report (BER) documentation, requires life-cycle maintenance of evaluations rather than one-off reviews, and introduces particulate risk management as a recurring requirement. Concurrently, the April 2025 Taryag Sterilization alliance is bringing sterilization data streams into EBI's testing workflow as an end-to-end service.

The Józefów laboratory is being expanded in physical capacity to absorb the EU MDR (European Medical Device Regulation) demand surge, with capital allocated to instrumentation and to recruiting quality and laboratory staff. Process Analytical Evaluation Threshold (AET) calculations, vendor-locked analytical data and air-gapped GLP (Good Laboratory Practice) workflows are the operational context that the expanded capacity will run into.

A single capability sits underneath each of those moving parts: process and quality data that can move from instrument to document to client without re-keying. That capability is what the analysis below maps onto.

Challenges we see

  • Operations Manufacturing

    Scaling manual workflows alongside a larger laboratory footprint

    The Józefów laboratory expansion increases physical capacity, but BEP and BER creation, Analytical Evaluation Threshold calculations and report assembly remain manual. As sample volume increases, labour overhead scales linearly with throughput rather than sub-linearly.

    Where report production scales one-to-one with sample volume, the marginal cost per test moves with the laboratory rather than against it, and lead time begins to track headcount.

  • Operations Operations

    Concentrating BEP authorship in a small group of senior scientists

    Biological Evaluation Plans and Reports are built around detailed review of device materials and manufacturing processes by named senior scientists, including the institute's CEO and Head of R&D. More than twenty years of prior study history sits in PDF reports that are not systematically searchable.

    When institutional knowledge of similar devices has to be reconstructed for each new plan, the time-to-first-draft of a BEP tracks the calendar availability of a handful of named individuals.

  • Digital Integration

    Working around vendor-locked analytical instrument data

    GC-MS, LC-MS and ICP-MS instruments from different manufacturers produce datasets in vendor-specific formats (Agilent, Waters, Shimadzu). AET calculation is a manual or semi-manual step that sits between instrument run and report, creating a known gap between run completion and result availability.

    Where analytical results live in separate vendor estates, longitudinal toxicological trend analysis depends on a manual reconciliation step that has to be re-done for each report.

  • Compliance Regulatory

    Meeting the expanded ISO 10993-1:2025 documentation burden

    The 2025 revision adds personnel competency evidence, clearer risk logic and life-cycle maintenance to BEP and BER documentation, and treats biological evaluation as a recurring activity rather than a one-off event. Particulate risk management is a new requirement on top.

    Where each evaluation now has to be maintained over the device life cycle, the documentation workload compounds with every product or process change, and the review queue begins to set the pace of release.

  • Digital Integration

    Operating within an air-gapped GLP laboratory

    GLP and FDA 21 CFR Part 58 (the US rule on electronic records and signatures in regulated environments) practices keep laboratory OT (Operational Technology, i.e. the instruments and controllers on the lab floor) disconnected from IT systems, and biocompatibility studies run for weeks or months without client visibility into study progress.

    Where instrument data stays on the instrument and study status is communicated manually, the available analytics on historical work are bounded by what can be reconstructed by hand.

Opportunities, by urgency and business impact

Each bubble is one opportunity, numbered to match the list below. Further right means it bites sooner; higher means a bigger effect on the business. A bigger bubble means a bigger implementation effort.

Source: A4BEE analysis of public sources

  1. Automating the Analytical Evaluation Threshold pipeline

    Scientists calculate the Analytical Evaluation Threshold from the device exposure duration and then configure chromatography software and interpret results against toxicological limits manually.

    A closed-loop pipeline that takes the Biological Evaluation Plan as input, sets AET parameters in the analytical instrument, ingests results from GC-MS and LC-MS, and flags peaks above the toxicological threshold for review cuts manual steps out of the characterisation run.

    • EBI Friction Matrix Analysis, January 2026
    • ISO 10993-1:2025 revision summary, EBI News

  2. Making twenty years of biocompatibility reports queryable

    The institute's biocompatibility testing history lives in thousands of PDF reports that cannot be searched systematically. Starting a BEP for a new silicone-based catheter means searching for prior studies on similar devices by hand.

    An on-premise retrieval layer with OCR (Optical Character Recognition) and semantic search over the institute's historical report corpus lets scientists surface every prior study on a similar device in seconds rather than over hours of manual reading.

    • EBI Technical Gap Analysis, January 2026
    • EBI About Us, https://ebi.bio/about-us/

  3. Bringing instrument results into one laboratory dashboard

    GC-MS and ICP-MS datasets are extracted manually from vendor software and pasted into reports. The interval between instrument run and final report is a known operational friction point that prevents real-time view of laboratory status.

    Vendor-agnostic connectors into the analytical instruments, writing results into one time-series model, make report assembly a query against current data rather than a manual compilation from per-run exports.

    • EBI Operational Friction Analysis, January 2026
    • EBI Deep Research Report, January 2026

  4. Giving clients real-time visibility into study progress

    Biocompatibility studies run for weeks or months. Project milestones such as 'sensitisation study initiated' or 'cytotoxicity results pending' are communicated manually rather than shown in a self-service view.

    A client-facing portal showing live study milestones and draft document availability turns study status into a service differentiator and removes the manual communications overhead that the project team carries today.

    • EBI Digital Friction Analysis, January 2026
    • EBI About Us, https://ebi.bio/about-us/

  5. Bridging IT and OT in the GLP laboratory

    GLP practice keeps laboratory instruments disconnected from the corporate network. Data transfer between instruments and reporting systems is done by hand, with the attendant risk of transcription error in records that have to satisfy FDA audits.

    VLAN (Virtual Local Area Network) segmentation and OPC UA (Open Platform Communications Unified Architecture) connectivity from instruments to the scientist's workstation move results as data, with an audit trail that holds up under 21 CFR Part 11 review.

    • EBI IT/OT Divide Analysis, January 2026
    • EBI FAQ, https://ebi.bio/faq/

What we'd propose

  • Digital Lab

    Automated Biological Evaluation Reporting Platform

    A Digital Lab solution that automates Biological Evaluation Report generation by integrating analytical instrument data into a compliant documentation workflow aligned with ISO 10993-1:2025.

    • Closed-loop Analytical Evaluation Threshold pipeline

      From plan parameter to flagged peak

      The Biological Evaluation Plan is the input; the platform sets the AET in the chromatography software, pulls results from GC-MS and LC-MS, and flags every peak above the toxicological threshold for immediate scientific review, removing the manual configuration step that today sits between run and report.

    • ISO 10993-1:2025 aligned report templates

      Documentation shaped for 2025

      Report templates carry the 2025 requirements as structured fields — personnel competencies, risk logic, life-cycle hooks and particulate risk — so the recurring documentation burden is built into the template rather than re-invented per study.

    • 21 CFR Part 11 audit trail

      Records that stand up in audit

      Each edit, electronic signature and result transfer is captured with timestamp and user identity, producing the evidence chain the FDA expects during a 21 CFR Part 11 inspection of a GLP study.

    • Report assembly moves from manual compilation to a query against current data.
    • Every calculation step leaves a trace the inspector can read directly.
    • The 2025 documentation burden is absorbed by the template rather than by the scientist.

    We've done this before

    See how we do Digital Lab
  • Enterprise AI

    Józefów Laboratory Data Backbone

    A vendor-agnostic data platform that connects Agilent, Waters and Shimadzu analytical instruments to a single time-series model, giving the laboratory one source of truth for toxicological trend analysis and real-time process intelligence.

    • Multi-vendor instrument integration

      All instruments writing to one model

      OPC UA connectors and instrument-specific drivers move results from Agilent, Waters and Shimadzu systems into a unified data lake without re-keying, so the analytical record is no longer split across vendor file formats.

    • Laboratory KPI dashboard

      Throughput and status in one view

      A real-time dashboard shows sample processing status, queue depth and throughput by analytical method, replacing manual extraction from per-run exports with a single screen that laboratory leads can read at a glance.

    • Historical trend and golden-run comparison

      Yesterday's best run is today's benchmark

      Current analytical runs are overlaid on historical best-run profiles so deviations are visible during the run rather than reconstructed from a report after the fact.

    • Vendor lock-in stops being a precondition for using the analytical data.
    • Laboratory leads see today's status without compiling it from per-run exports.
    • Longitudinal toxicological trend analysis becomes a dashboard, not a project.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Secure IT/OT Convergence for the GLP laboratory

    A network architecture with VLAN segmentation and OPC UA connectivity that bridges the air gap between laboratory instruments and enterprise systems while keeping the regulatory posture of a GLP environment.

    • VLAN segmentation with defined gateways

      OT zones, controlled access

      The network topology separates laboratory instrument traffic from corporate systems through defined gateways, so data flows from instrument to scientist without exposing equipment to the open internet.

    • OPC UA connectivity layer

      Standard protocol on legacy instruments

      OPC UA servers on the analytical instruments expose results in a documented, vendor-neutral form, letting downstream systems consume data through a single protocol rather than per-vendor integrations.

    • 21 CFR Part 11 electronic records baseline

      Audit-grade by construction

      Identity management, electronic signatures and audit-trail handling are built into the data path so records move from instrument to report with the provenance required for a 21 CFR Part 11 review.

    • Manual data transfer between instrument and report is removed, and so is its transcription-error risk.
    • The laboratory's regulatory posture is preserved as instruments become networked.
    • Downstream analytics can build on instrument data without re-keying.

    We've done this before

    See how we do Digital Lab
  • Enterprise AI

    AI-assisted retrieval over historical biocompatibility studies

    A private, on-premise retrieval system that uses optical character recognition and semantic search to make the institute's twenty-plus years of biocompatibility reports queryable as a knowledge corpus for accelerating BEP authorship.

    • OCR and semantic extraction from legacy reports

      PDFs become a corpus

      Optical character recognition and semantic tagging convert the institute's historical PDF reports into a structured corpus indexed by device type, material, test outcome and regulatory finding.

    • Similar-device retrieval at BEP authorship

      Prior studies surface in seconds

      When a scientist starts a BEP for a new silicone-based catheter, the retrieval layer returns every prior study on similar devices with the relevant toxicological and biological evaluation findings highlighted.

    • Risk-logic drafting support for ISO 10993-1:2025

      The risk-based approach gets a head start

      The retrieval layer surfaces the historical data that supports the risk-based reasoning the 2025 standard expects in a BEP, so the scientist edits and judges rather than reconstructs the prior art from scratch.

    • The first draft of a BEP starts from the institute's own prior art rather than from a blank page.
    • Institutional knowledge stops being bound to the calendars of named senior scientists.
    • The 2025 risk-based approach is supported by a retrieval layer rather than by memory.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Client study visibility portal

    A client-facing portal that gives medical device manufacturers real-time visibility into the progress of their biocompatibility studies, including milestone status and draft document availability, turning study status into a service differentiator.

    • Milestone tracking interface

      Study phase at a glance

      An interactive timeline shows study initiation, each testing phase and expected completion dates, with status updated as the laboratory moves through the protocol.

    • Secure document review

      Draft BERs delivered as a workflow

      Authenticated client access to draft Biological Evaluation Reports and preliminary results, with commenting captured in the same audit trail the laboratory keeps internally.

    • Notification engine

      Status updates without manual email

      Automated alerts for milestone completions, result availability and timeline changes, removing the manual communication overhead that the project team carries today.

    • Study status becomes part of the service rather than an email the team sends.
    • Client review of draft BERs moves into the same audit trail as the study itself.
    • Differentiator on transparency in a CRO market where lead time is the headline metric.

    We've done this before

    See how we do Digital Lab

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what European Biomedical Institute (EBI)'s own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Instrument data integration 35 → 80
Analytical instrument data lives in vendor-specific formats from Agilent, Waters and Shimadzu systems and is reconciled manually for each report. A unified data layer changes what is available for trend analysis without changing what the instruments measure.
Process automation 30 → 75
BEP and BER creation, Analytical Evaluation Threshold calculation and report assembly are scientist-driven workflows. Template and pipeline automation removes the parts that scale linearly with sample volume.
IT/OT convergence 25 → 70
GLP practice keeps laboratory instruments disconnected from corporate networks, with manual data transfer as the bridge. VLAN segmentation and OPC UA close the gap without compromising the regulatory posture.
Knowledge retrieval 20 → 70
Twenty-plus years of biocompatibility study history is held in PDF reports that are not systematically searchable. Retrieval over that corpus changes how quickly prior studies surface during BEP authorship.
Client experience 40 → 85
Clients value execution speed but communicate manually with the project team during multi-week studies. A self-service visibility portal moves status into the service itself.
Regulatory technology 45 → 80
ISO 17025 and GLP certification are in place; the 2025 ISO 10993-1 revision expands the documentation burden from a one-off event to a life-cycle activity. Template and agent-assisted drafting absorb the recurring work.

Check this yourself

Our Service Portal has free self-assessments and market comparisons. These are the ones that line up with what we've read above — no sales call required.

Think we've read this right?

Talk to us

Related reading

This is an independent analysis prepared by A4BEE from publicly available information as of January 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with European Biomedical Institute (EBI), and may be incomplete or inaccurate. All company names and trademarks are the property of their respective owners. To request a correction or removal, contact [email protected].