Projektbeschreibung
Raman-Spektroskopie verbessert Biotherapeutika
Biotherapeutika sind der am schnellsten wachsende Bereich der pharmazeutischen Industrie. Doch die Branche kämpft mit kritischen Anforderungen an geringere Kosten, effizientere und schnellere Bioverarbeitung sowie stabile Ergebnisse. Die Raman-Spektroskopie ist eine Technologie, die die Bioverarbeitung verbessern könnte. Im EU-finanzierten Projekt SpectraHow wird eine skalierbare Software auf Basis innovativer Techniken des maschinellen Lernens entwickelt, die Daten erarbeitet, automatisch die Prozessdynamik misst und Informationen überträgt. Das Projekt will die Verwendung der Software noch verbessern, Muster und Algorithmen optimieren und ihre Kompatibilität mit anderen Software-Instrumenten und spektrologischen Technologien nachweisen.
Ziel
Today, biopharmaceuticals generate global revenues of €145 billion, representing about 20% of the pharma market. The growth rate of biopharma, 8%, currently doubles that of conventional pharma. However, biopharma companies are facing an increased pressure for cheaper process development, faster time to market as well as more consistent production (reducing failure). Raman spectroscopy is a promising technology to improve bio processing and manufacture. Nevertheless, the corresponding data analysis of the complex spectra in biopharma is more challenging that conventional pharma, taking 2-5 months in order to acquire enough data and train a reliable predictive model. Founded in 2017, DataHow is a spin-off from the ETH Zürich, with the aim to accelerate biopharmaceutical process development and reduce risks in production through advanced algorithm-enabled digital solutions; SpectraHow is our scalable software based on advanced machine learning techniques, capable of accurately handling the large amounts of data generated, to automatically quantify the process dynamics and translate the information into direct decision support. Our algorithmic toolbox at TRL6 has been already tested in operational environment for different processes with several pharma companies, with a precision of up to 35% more accurate compared to commercial tools, and a smaller effort due to our automated calibration procedure. Next steps to take forward SpectraHow to commercialization are: improve usability with a Graphical User Interface, fine-tune our models and algorithms to further automate analysis, and show compatibility of our solution with third-party software tools to expand its functionality from R&D to production and to other spectroscopy techniques. In the booming Raman for pharma market valued at 2.5 billion in 2015, we forecast generate €21 Million in revenues in 2026 while creating 14 new jobs.
Wissenschaftliches Gebiet
- natural sciencescomputer and information sciencessoftware
- engineering and technologyenvironmental biotechnologybioremediationbioreactors
- natural sciencescomputer and information sciencesdata sciencebig data
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensors
- natural sciencesphysical sciencesopticsspectroscopy
Programm/Programme
Thema/Themen
Aufforderung zur Vorschlagseinreichung
Andere Projekte für diesen Aufruf anzeigenUnterauftrag
H2020-SMEInst-2018-2020-1
Finanzierungsplan
SME-1 - SME instrument phase 1Koordinator
8093 Zurich
Schweiz
Die Organisation definierte sich zum Zeitpunkt der Unterzeichnung der Finanzhilfevereinbarung selbst als KMU (Kleine und mittlere Unternehmen).