Spacemaker
Spacemaker
Spacemaker
● The question is :
How will already-crowded urban areas fit all those extra people?
That’s what Spacemaker is trying to figure out.
● Spacemaker is a company using AI software to simulate and optimize how buildings and
structures would fit in a building site.
● The company’s technology lets engineers, real estate developers and architects go
through multiple site proposals and sort out the best ones.
● Spacemaker CEO and co-founder Håvard Haukeland said in a statement.
“With Spacemaker they can, for the first time, generate and assess billions of
possible solutions to multi-building residential development design in only
hours."
According to the official company register Spacemaker AI was founded in late 2016 by Håvard Haukeland
and Carl Christensen. Håvard Haukeland had an educational background as an architect and within
technology having worked as an architect and within urban design. Carl Christensen had an educational
background as a computer engineer with an additional master’s in business administration and
economics.
In late 2017 there is an article that states their digital solution was part of a 4-year scientific research
project with support from amongst others Aspelin Ramm, AF Eiendom, Stor-Oslo Eiendom and SINTEF.
Spacemaker AI had then received upwards of 15 million Norwegian kroners ($1,8 million) from The
Research Council of Noway, Innovation Norway and Simula Research Laboratory to develop their digital
solution.
In October 2017 they had in one year grown into a small company with 20 employees. They were at the
time talking about using artificial intelligence in city planning and for building sites. In September 2017 a
group of investors came together to invest 22 million Norwegian kroners ($2,65 million) in the company. In
addition to this part of the Norwegian leadership group in CBRE decided to invest with CBRE being one of
the largest consultants in commercial property registered in New York. This is what part of their solution
looked like in late 2017: Only two weeks after their first project they claimed to have been approached by
more than 100 clients for collaboration and Håvard Haukeland was nominated as top leadership talent of
the year in Norway.
Designers draw to think of space, draw to explore the solution of spatial configurations, and
draw to make space. While SpaceMaker supports a convenient paper-like modeling
environment, there are still some limitations due to the computational implementation. For
example, the recognition systems currently only support single-stroke recognition. Users must
complete any single task, such as enclosing a building boundary, partitioning the floor, and
inserting a symbol, within only one single stroke input. In addition, SpaceMaker currently only
recognizes a building boundary with a simple rectangular shape. It is obvious that many
opportunities for extending the scope of this study remain. In the immediate future, usability
tests or other formative evaluation techniques such as heuristic evaluation should be used to
further refine the prototype. In addition, future versions of SpaceMaker should incorporate non-
rectilinear recognition, multi-stroke input, high-level context recognition, and multi-floor layout.
SpaceMaker represents our intention to look at a different direction for establishing a 3-D
modeling environment that is truly suitable to the early conceptual development for the
architectural design.
Space makes combines expertise from a wide range of fields including architecture,
mathematics, physices, machine learning and optimization. We provide a user with creative set
of high quality site proposals. By applying machine learning with reinforcing mechanisms across
our system, we improve our optimization algorithms after every run.