Predictive Modelling to Calculate the Need for Office Space
Leppänen, Petri (2024)
Leppänen, Petri
2024
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-202404096114
https://urn.fi/URN:NBN:fi:amk-202404096114
Tiivistelmä
Remote work or hybrid work after the pandemic is the new default in people's
workplaces and many have been making good home offices themselves. This is
why increasingly many offices are empty most of the time. This kind of prediction
can be a good tool to help seek almost the right size offices and with presence
sensors and reserved spaces; there can be enough space for different type of
work. Meeting rooms, quiet rooms or areas, and ad hoc spaces just to sit down
and print out papers and move to a meeting are the kinds of places employees
need when they come to the office. For example, lounge and café type of spaces
are also good for small talk with persons outside of your team. The purpose of an
office space has changed after the pandemic and the company needs to be ready
to act based on different kinds of variables and predictions.
In this thesis, the focus was on investigating the possibility of making some kind
of prediction that can be used to anticipate office sizes in the future. My employer
asked me to see if there is some kind of library or model to predict our office
places in the future. This project-based thesis aimed to make the first iteration of
a prediction that can be used and modified according to company needs. The
commissioning company provided all the data needed for this thesis work.
The results show that the prediction met the requirements of the commission.
With more data, the prediction will be more accurate. Prophet may not be the best
for doing predictions and therefore the suggestion is to further investigate differ ent libraries for different kinds of needs. Longer periods of data will get better
prediction results and also adjusting variables that are used to make predictions
would be ones that need more investigating.
workplaces and many have been making good home offices themselves. This is
why increasingly many offices are empty most of the time. This kind of prediction
can be a good tool to help seek almost the right size offices and with presence
sensors and reserved spaces; there can be enough space for different type of
work. Meeting rooms, quiet rooms or areas, and ad hoc spaces just to sit down
and print out papers and move to a meeting are the kinds of places employees
need when they come to the office. For example, lounge and café type of spaces
are also good for small talk with persons outside of your team. The purpose of an
office space has changed after the pandemic and the company needs to be ready
to act based on different kinds of variables and predictions.
In this thesis, the focus was on investigating the possibility of making some kind
of prediction that can be used to anticipate office sizes in the future. My employer
asked me to see if there is some kind of library or model to predict our office
places in the future. This project-based thesis aimed to make the first iteration of
a prediction that can be used and modified according to company needs. The
commissioning company provided all the data needed for this thesis work.
The results show that the prediction met the requirements of the commission.
With more data, the prediction will be more accurate. Prophet may not be the best
for doing predictions and therefore the suggestion is to further investigate differ ent libraries for different kinds of needs. Longer periods of data will get better
prediction results and also adjusting variables that are used to make predictions
would be ones that need more investigating.