Loss-in-Weight feeder performance prediction using Machine Learning

Hikaru Graeme Jolliffe; Carlota Mendez Torrecillas; Gavin K. Reynolds; Richard Elkes; Hugh Verrier; Michael Devlin; Bastiaan H.J. Dickhoff; John Robertson

doi:10.1016/B978-0-443-28824-1.50016-8

Loss-in-Weight feeder performance prediction using Machine Learning

Hikaru Graeme Jolliffe^*, Carlota Mendez Torrecillas, Gavin K. Reynolds, Richard Elkes, Hugh Verrier, Michael Devlin, Bastiaan H.J. Dickhoff, John Robertson

^*Corresponding author for this work

Strathclyde Institute Of Pharmacy And Biomedical Sciences

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

Abstract

There has been significant drive in recent years to gain greater understanding of unit operations useful for continuous direct compaction, and to leverage data-driven approaches such as Machine Learning to extract trends from large and complex datasets. In this work, an approach using three Machine Learning models to predict the parameters in an equation for Loss-in-Weight feeder performance are presented. Industrially-relevant feeders with multiple screws per feeder are studied, and the approach allows feed factor decay to be predicted using material properties and equipment choice as inputs. Using a wide range of excipients and Active Pharmaceutical Ingredients (APIs) for testing shows good performance against industrially-relevant targets. The approach presented here would be useful for equipment pre-selection activities prior to experimental work.

Original language	English
Title of host publication	Computer Aided Chemical Engineering
Subtitle of host publication	Volume 53: 34th European Symposium on Computer Aided Process Engineering / 15th International Symposium on Process Systems Engineering
Pages	91-96
Volume	53
ISBN (Electronic)	978-0-443-28824-1
DOIs	https://doi.org/10.1016/B978-0-443-28824-1.50016-8
Publication status	Published - 26 Jun 2024
Event	European Symposium on Computer Aided Process Engineering and International Symposium on Process Systems Engineering - Florence, Italy Duration: 2 Jun 2024 → 6 Jun 2024 https://www.aidic.it/escape34-pse24/

Conference

Conference	European Symposium on Computer Aided Process Engineering and International Symposium on Process Systems Engineering
Abbreviated title	ESCAPE34-PSE24
Country/Territory	Italy
City	Florence
Period	2/06/24 → 6/06/24
Internet address	https://www.aidic.it/escape34-pse24/

Funding

This work has been funded by the Medicines Manufacturing Innovation Centre project (MMIC), UK (project ownership: Centre for Process Innovation, CPI). Funding has come from Innovate UK and Scottish Enterprise. Founding industry partners with significant financial and technical support are AstraZeneca and GSK. The University of Strathclyde (via CMAC) is the founding academic partner. Pfizer are project partners and have provided key technical input and data. Project partners DFE Pharma have provided materials and technical input, and project partners Gericke AG have provided technical equipment support and advice. Project partners Siemens and Applied Materials have provided key software and software/IT expertise. Andrew Shier (now of GSK) significantly contributed to the present work via the literature data

Keywords

feeder
modelling
machine learning
powders

Access to Document

10.1016/B978-0-443-28824-1.50016-8

1 Finished

GC1: MMIC Grand Challenge 1
Robertson, J. (Principal Investigator)
2/07/18 → 13/08/21
Project: Research

1 Poster

Medicines Manufacturing Innovation Centre (MMIC): the use of a Digital Twin
Jolliffe, H., Mendez Torrecillas, C., Prostredny, M., Devlin, M., Bordos, E., Markl, D. & Robertson, J., 16 May 2022, p. 9-9. 1 p.
Research output: Contribution to conference › Poster

Open Access
File

Cite this

Jolliffe, H. G., Mendez Torrecillas, C., Reynolds, G. K., Elkes, R., Verrier, H., Devlin, M., Dickhoff, B. H. J., & Robertson, J. (2024). Loss-in-Weight feeder performance prediction using Machine Learning. In Computer Aided Chemical Engineering: Volume 53: 34th European Symposium on Computer Aided Process Engineering / 15th International Symposium on Process Systems Engineering (Vol. 53, pp. 91-96) https://doi.org/10.1016/B978-0-443-28824-1.50016-8

@inproceedings{6566d17b9c914f6fbdeae01b322e27e4,

title = "Loss-in-Weight feeder performance prediction using Machine Learning",

abstract = "There has been significant drive in recent years to gain greater understanding of unit operations useful for continuous direct compaction, and to leverage data-driven approaches such as Machine Learning to extract trends from large and complex datasets. In this work, an approach using three Machine Learning models to predict the parameters in an equation for Loss-in-Weight feeder performance are presented. Industrially-relevant feeders with multiple screws per feeder are studied, and the approach allows feed factor decay to be predicted using material properties and equipment choice as inputs. Using a wide range of excipients and Active Pharmaceutical Ingredients (APIs) for testing shows good performance against industrially-relevant targets. The approach presented here would be useful for equipment pre-selection activities prior to experimental work.",

keywords = "feeder, modelling, machine learning, powders",

author = "Jolliffe, {Hikaru Graeme} and {Mendez Torrecillas}, Carlota and Reynolds, {Gavin K.} and Richard Elkes and Hugh Verrier and Michael Devlin and Dickhoff, {Bastiaan H.J.} and John Robertson",

year = "2024",

month = jun,

day = "26",

doi = "10.1016/B978-0-443-28824-1.50016-8",

language = "English",

volume = "53",

pages = "91--96",

booktitle = "Computer Aided Chemical Engineering",

note = "European Symposium on Computer Aided Process Engineering and International Symposium on Process Systems Engineering, ESCAPE34-PSE24 ; Conference date: 02-06-2024 Through 06-06-2024",

url = "https://www.aidic.it/escape34-pse24/",

}

Jolliffe, HG , Mendez Torrecillas, C, Reynolds, GK, Elkes, R, Verrier, H, Devlin, M, Dickhoff, BHJ & Robertson, J 2024, Loss-in-Weight feeder performance prediction using Machine Learning. in Computer Aided Chemical Engineering: Volume 53: 34th European Symposium on Computer Aided Process Engineering / 15th International Symposium on Process Systems Engineering. vol. 53, pp. 91-96, European Symposium on Computer Aided Process Engineering and International Symposium on Process Systems Engineering, Florence, Italy, 2/06/24. https://doi.org/10.1016/B978-0-443-28824-1.50016-8

Loss-in-Weight feeder performance prediction using Machine Learning. / Jolliffe, Hikaru Graeme ; Mendez Torrecillas, Carlota; Reynolds, Gavin K. et al.
Computer Aided Chemical Engineering: Volume 53: 34th European Symposium on Computer Aided Process Engineering / 15th International Symposium on Process Systems Engineering. Vol. 53 2024. p. 91-96.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution book

TY - GEN

T1 - Loss-in-Weight feeder performance prediction using Machine Learning

AU - Jolliffe, Hikaru Graeme

AU - Mendez Torrecillas, Carlota

AU - Reynolds, Gavin K.

AU - Elkes, Richard

AU - Verrier, Hugh

AU - Devlin, Michael

AU - Dickhoff, Bastiaan H.J.

AU - Robertson, John

PY - 2024/6/26

Y1 - 2024/6/26

N2 - There has been significant drive in recent years to gain greater understanding of unit operations useful for continuous direct compaction, and to leverage data-driven approaches such as Machine Learning to extract trends from large and complex datasets. In this work, an approach using three Machine Learning models to predict the parameters in an equation for Loss-in-Weight feeder performance are presented. Industrially-relevant feeders with multiple screws per feeder are studied, and the approach allows feed factor decay to be predicted using material properties and equipment choice as inputs. Using a wide range of excipients and Active Pharmaceutical Ingredients (APIs) for testing shows good performance against industrially-relevant targets. The approach presented here would be useful for equipment pre-selection activities prior to experimental work.

AB - There has been significant drive in recent years to gain greater understanding of unit operations useful for continuous direct compaction, and to leverage data-driven approaches such as Machine Learning to extract trends from large and complex datasets. In this work, an approach using three Machine Learning models to predict the parameters in an equation for Loss-in-Weight feeder performance are presented. Industrially-relevant feeders with multiple screws per feeder are studied, and the approach allows feed factor decay to be predicted using material properties and equipment choice as inputs. Using a wide range of excipients and Active Pharmaceutical Ingredients (APIs) for testing shows good performance against industrially-relevant targets. The approach presented here would be useful for equipment pre-selection activities prior to experimental work.

KW - feeder

KW - modelling

KW - machine learning

KW - powders

U2 - 10.1016/B978-0-443-28824-1.50016-8

DO - 10.1016/B978-0-443-28824-1.50016-8

M3 - Conference contribution book

VL - 53

SP - 91

EP - 96

BT - Computer Aided Chemical Engineering

T2 - European Symposium on Computer Aided Process Engineering and International Symposium on Process Systems Engineering

Y2 - 2 June 2024 through 6 June 2024

ER -

Jolliffe HG , Mendez Torrecillas C, Reynolds GK, Elkes R, Verrier H, Devlin M et al. Loss-in-Weight feeder performance prediction using Machine Learning. In Computer Aided Chemical Engineering: Volume 53: 34th European Symposium on Computer Aided Process Engineering / 15th International Symposium on Process Systems Engineering. Vol. 53. 2024. p. 91-96 doi: 10.1016/B978-0-443-28824-1.50016-8

Loss-in-Weight feeder performance prediction using Machine Learning

Abstract

Conference

Funding

Keywords

Access to Document

Fingerprint

Projects

GC1: MMIC Grand Challenge 1

Research output

Medicines Manufacturing Innovation Centre (MMIC): the use of a Digital Twin

Cite this