☰

EcoBean

From coffee waste to five bio-based ingredients

Industry
Biotechnology
Headquarters
Warsaw, Poland
Public information as of
January 2026

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

Strategic priorities

EcoBean is a Warsaw University of Technology spin-off that valorizes the 12 million tonnes of spent coffee grounds produced globally each year into five bio-based fractions: coffee oil, antioxidants, polylactic acid (PLA), lignin, and protein additives. In 2024 the Polish Agency for Enterprise Development (PARP) awarded the company a nearly €7 million grant under the FENG programme to build the EcoBean Technology Center (ETC), described in the grant announcement as an 'intelligent and fully sustainable demo factory' targeted at 1,000 tonnes of feedstock per year and built around a Digital Twin.

The defining property of the feedstock sets the operational shape of the business. Spent coffee grounds carry roughly 60 percent moisture and begin to mold within 72 hours, which compresses the logistics window around receipt and processing. A single instant-coffee factory can produce up to 40,000 tonnes of spent grounds a year, and EcoBean's 'Plug & Play' concept installs modular biorefinery units at customer sites so that processing happens where the feedstock is generated.

Moving the chemistry from more than 8,000 laboratory experiments to a continuous industrial line is the engineering step that the ETC exists to validate. The Chief Digital Officer has framed the Digital Twin as 'an element of mitigating technology scale-up risk', enabling what-if analyses against live operational boundaries. The same data path also has to support a distributed fleet of customer-site modules monitored remotely from Warsaw and a batch-level digital product passport for the high-margin cosmetics and pharmaceutical buyers.

The company is also a recent MassChallenge Switzerland Climate Resilience Prize laureate (CHF 100,000 in 2024) and counts EIT InnoEnergy among its institutional investors, alongside a 2022 seed round of about $2.18M led by CofounderZone, CIECH Ventures and COBIN Angels. The combination of non-dilutive grant funding, an investor base connected to large industrial energy players, and a CDO-led digital agenda puts the next eighteen months at the ETC firmly inside the question of how data moves between simulation, lab, line, and the customer's site.

Challenges we see

  • Operations Manufacturing

    Processing feedstock within its biological window

    Spent coffee grounds carry roughly 60 percent moisture and begin to mold within 72 hours of generation, which sets a hard upper bound on the time between collection and processing that traditional waste-collection systems are not designed for.

    Where the route from collection to reactor is hours rather than days, the inventory in front of the line and the consistency of the feedstock arriving at it become part of the process specification rather than inputs the line tolerates.

  • Digital Integration

    Connecting simulation models to physical process control

    EcoBean has committed to an INOSIM Digital Twin for the ETC. The simulation platform is procured; the physical line, its programmable logic controllers (PLCs) and supervisory control layer are being built in parallel at the same site.

    The value of a Digital Twin depends on whether the simulation receives live process data and feeds back into operating decisions, which means the boundary between INOSIM and the line's control layer has to be designed before the line is commissioned.

  • Digital Operations

    Bringing lab quality results into the production record

    The Nadarzyn quality-control laboratory uses OHAUS balances and analytical instruments to characterize feedstock and product fractions; results are the basis for both batch release and ongoing process development.

    Where laboratory results are read off instruments and re-keyed into other systems, each transfer is its own error source and its own audit-trail gap, so the time from sample to a production-side decision grows in proportion to the number of manual steps.

  • Operations Manufacturing

    Operating modular units from a central fleet view

    The Plug & Play model places biorefinery modules at customer sites such as Delta and PRIO, where local staff operate the unit and EcoBean staff diagnose and optimize from Warsaw.

    When a unit's operating picture lives only on the unit, every intervention requires physical presence, which makes engineering hours, customer downtime and the per-unit cost of ownership all move in the same direction.

  • Compliance Regulatory

    Protecting proprietary process knowledge across deployments

    EcoBean's competitive position rests on proprietary extraction protocols that travel with the deployed module into a customer environment whose network, staff and security posture are outside EcoBean's direct control.

    When a process recipe is executed on equipment connected to an unfamiliar network, the security boundary around the recipe has to be built into the deployment rather than added on top of it.

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. Closing the loop between INOSIM and the ETC control layer

    Moving from more than 8,000 laboratory experiments to a continuous industrial process introduces risks around heat transfer, mass balance and mechanical behaviour that physical pilot runs cannot exhaustively test.

    A live bridge between the INOSIM simulation engine and the line's PLC and supervisory control and data acquisition (SCADA) layer lets the operations team run parametric what-if scenarios against the current operating envelope and compare the prediction to the running line.

    • INOSIM case study, From Waste to Value — Simulation for EcoBean's Sustainable Solution, 2026
    • PARP FENG grant announcement, EcoBean Technology Center, EIT Food news, 2024

  2. Feeding laboratory results into the biorefinery control loop

    Spent coffee grounds from different origins vary in moisture, oil content and contamination, and the current information path from the Nadarzyn laboratory to the production line goes through spreadsheets rather than the control system.

    Connecting balances and analytical instruments to a shared data layer lets feedstock characterization move from the laboratory into the production record with its own provenance, and lets downstream parameters be adjusted against measured feedstock properties.

    • OHAUS, The story of EcoBean: Brewing sustainable innovation from coffee grounds, 2025
    • EcoBean, About Us (ecobean.pl/about-us)

  3. Building the remote-operations picture before the fleet grows

    Plug & Play modules will be installed at customer sites, and EcoBean intends to supervise yield, health and energy use from Warsaw, but the central operations view is not yet built and each new site currently extends the work.

    A cloud-hosted fleet layer designed before the first customer-site installation establishes the data model, the alarm taxonomy and the role-based access rules that subsequent sites will inherit, rather than accumulating as one-offs.

    • EcoBean, Projects (ecobean.pl/projects)
    • EIT Food impact story, EcoBean: turning coffee waste into valuable chemicals, 2025

  4. Producing batch-level evidence for cosmetics and pharmaceutical buyers

    Antioxidants and other fractions sold into cosmetics and pharmaceutical markets require batch-level documentation covering origin, purity and processing conditions, and the FENG grant requires per-batch reporting of energy and water use.

    An automated evidence pipeline ties feedstock intake, processing conditions, laboratory results and energy data to a per-batch record so that the documentation required by the buyer and by the funder is generated rather than assembled.

    • EcoBean, Impact (ecobean.pl/impact)
    • EcoBean press release, EcoBean with nearly €7 million grant from PARP, EIT Food, 2024

  5. Setting the security baseline for distributed deployments

    Proprietary extraction protocols will run on equipment deployed in customer environments whose networks and staff EcoBean does not control, and the current security perimeter is the headquarters building.

    A Zero Trust architecture for the deployed unit, designed to IEC 62443, treats the customer's network as untrusted by default and contains the recipe inside an encrypted, role-controlled boundary that travels with the unit.

    • EcoBean, Projects (ecobean.pl/projects)
    • A4BEE article, Cybersecurity for industrial automation and control systems in life sciences

What we'd propose

  • Digital CDMO

    Digital Twin integration architecture for the ETC

    We connect the INOSIM simulation platform to the ETC's PLC and SCADA layer through a documented interface, so the running line feeds the simulation and the simulation can be queried against the live operating state.

    • Live OPC UA bridge

      From line to simulation, in a documented form

      Publish process values from the line's controllers and sensors through OPC UA (Open Platform Communications Unified Architecture) so INOSIM and the supervisory layer share one documented, vendor-neutral data path instead of point-to-point scripts.

    • What-if scenarios against the running line

      Test changes before they reach the line

      Run parametric scenarios inside INOSIM using the current operating envelope as the baseline, compare the prediction to the running process, and record the result for review before any physical change is made.

    • Operating-state dashboard for engineers

      Live process picture, not a retrospective report

      Build a Grafana-based view that shows the live process state against the simulation's prediction, with deviations highlighted, so the engineering team reads the line in the same shape they read the model.

    • Scale-up risk is read against a live model rather than against a sequence of physical pilot runs.
    • The simulation stays current with the line, so what-if analyses stop being a one-off exercise.
    • What the line did and what the model said can be compared on the same screen.

    We've done this before

    See how we do Digital CDMO
  • Digital Lab

    Integrated lab-to-production data path for the Nadarzyn QC lab

    We connect the Nadarzyn laboratory's balances and analytical instruments to a shared data layer that the biorefinery control system can read, so quality results reach the line as data with their own provenance.

    • Instrument integration

      Results captured at the instrument

      Connect OHAUS balances and analytical instruments so each result is captured with instrument identity, method version and timestamp attached, rather than read from a screen and re-entered.

    • Sample and batch lineage

      From feedstock to finished fraction, traced

      Link each sample to the feedstock lot and to the downstream batch record, so a quality result can be followed in both directions from any point in the chain.

    • Quality-triggered process parameter adjustment

      Feedstock variability handled in the control loop

      Surface measured feedstock properties to the biorefinery control layer so that operating parameters can be adjusted against measured input quality rather than against a fixed recipe.

    • Re-keying and transcription are removed from the path between laboratory and line.
    • Feedstock variability is handled in the control loop, not by operator judgement.
    • Each batch carries a complete quality history that can be reviewed end to end.

    We've done this before

    See how we do Digital Lab
  • Enterprise AI

    Secure remote-operations platform for the Plug & Play fleet

    We build a cloud-hosted fleet view that lets the Warsaw team see the operating state of every deployed module, with the security boundary designed so the customer's network does not need to be trusted.

    • Centralized fleet dashboard

      Every module on one screen

      Aggregate yield, energy use and alarm state from every deployed module into a single operational view, with drill-down to the per-unit detail and the historical record.

    • Zero Trust edge connectivity

      IEC 62443 zones from the first deployment

      Establish network segmentation and identity-based access controls between the customer site and EcoBean's operations environment to IEC 62443, so the customer's network is treated as untrusted by default.

    • Remote diagnostics and guided maintenance

      Engineers reach the unit without travelling

      Provide the Warsaw engineering team with the tooling to diagnose, configure and walk a customer-site operator through routine interventions without physical travel.

    • New sites inherit the security and operating model rather than reinventing it.
    • Engineering hours per unit fall as the fleet grows.
    • Customer downtime drops because most interventions are remote.

    We've done this before

    See how we do Enterprise AI
  • Enterprise AI

    Automated ESG and grant-compliance data pipeline

    We build the data pipeline that turns the ETC's operating data into the documentation required by the FENG grant and by customers asking for sustainability evidence.

    • Sensor-to-report data capture

      Energy, water and emissions measured at source

      Capture energy, water and emissions data directly from the process equipment, validate it against expected ranges, and write it into a structured per-batch record.

    • Digital product passport generator

      Every batch carries its origin and footprint

      Generate a per-batch digital product passport that links feedstock origin, processing conditions, laboratory results and the calculated carbon footprint, in the format a cosmetics or pharmaceutical buyer can use.

    • PARP and FENG reporting view

      Grant reporting as a by-product of operations

      Build a reporting view that pulls from the same operational data store, so FENG programme submissions are produced from the system rather than assembled by hand.

    • Sustainability and grant documentation is generated rather than compiled.
    • Buyers receive a consistent product passport across batches and customers.
    • Executive time spent on compliance reporting falls in proportion to the automation.

    We've done this before

    See how we do Enterprise AI
  • Agents

    AI agents for compliance documentation and product passports

    We deploy narrow, reviewable agents that draft the recurring parts of EcoBean's documentation work — FENG grant submissions, digital product passports, customer sustainability responses — from the underlying records, with a named reviewer approving every output.

    • Drafting from source records

      First drafts from operational data

      Generate the first draft of a FENG progress submission, a digital product passport, or a customer sustainability response directly from the per-batch operational and laboratory records, so the author edits and judges rather than assembles.

    • Template and completeness checking

      Gaps found before review

      Check a draft submission against the grant template or a buyer's passport specification, and return missing or inconsistent sections before the document enters the human review queue.

    • Cross-document consistency search

      One number, many documents

      When a figure, a process description or a feedstock specification changes, retrieve every controlled document that references it and rank them by how directly they are affected, so the update scope is known on day one.

    • Documentation cycles shrink because drafts arrive complete and on template.
    • Buyers and grant reviewers see the same evidence, drawn from the same source.
    • Every output is traceable to the records it came from and signed off by a named reviewer.

    We've done this before

    See how we do Agents

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what EcoBean's own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Digital Twin maturity 35 → 80
INOSIM has been procured for the ETC, but the simulation is not yet wired to the physical line; the target requires the simulation to read live process state and to be queried against the running line.
Lab digitization 25 → 75
Laboratory data is captured on OHAUS instruments and recorded manually, and the Nadarzyn lab is not connected to the production control layer; the target requires instrument-level capture with downstream use of the result.
IT/OT convergence 20 → 85
The Plug & Play unit concept is published but the central fleet view and the remote-operations tooling are not yet in place; the target requires a single operating picture across headquarters, the ETC and customer-site modules.
Data platform integration 30 → 80
Data lives in the laboratory, in process equipment and in the simulation, each in its own format; the target requires a shared data layer that the laboratory, the line and the simulation can all read and write against.
Cybersecurity posture 25 → 75
Security is currently anchored at the headquarters perimeter; the target requires a deployment-ready architecture that treats customer-site networks as untrusted and protects proprietary process recipes in transit and at rest.
ESG and grant automation 15 → 70
FENG grant reporting and digital product passports are assembled manually from operational records; the target requires documentation generated as a by-product of running the line.

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 EcoBean, 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].