See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/291696318

DAYLIGHTING IN HOSPITAL PATIENT ROOMS:

PARAMETRIC WORKFLOW AND GENETIC
ALGORITHMS FOR AN...

Conference Paper December 2015

CITATION READS

1 507

4 authors, including:

Ahmed Sherif Hanan Sabry

The American University in Cairo Ain Shams University
27 PUBLICATIONS 163 CITATIONS 19 PUBLICATIONS 90 CITATIONS

SEE PROFILE SEE PROFILE

Ayman Wagdy
Queensland University of Technology
19 PUBLICATIONS 44 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Kaleidocycle Origami facade View project

All content following this page was uploaded by Ayman Wagdy on 24 January 2016.

The user has requested enhancement of the downloaded file.

Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.
DAYLIGHTING IN HOSPITAL PATIENT ROOMS: PARAMETRIC WORKFLOW
AND GENETIC ALGORITHMS FOR AN OPTIMUM FAADE DESIGN
Ahmed Sherif1, Hanan Sabry2, Ayman Wagdy1, Rasha Arafa1
1
The American University in Cairo (AUC), Cairo, Egypt
2
Faculty of Engineering, Ain Shams University, Cairo, Egypt
contact email: asherif@aucegypt.edu
ABSTRACT
Daylighting and access to external view in hospital
patient rooms can positively contribute to patients
healing and improve staff performance.
This paper introduced a parametric workflow and
optimization for generating and evaluating alternative
faade configurations. The approach developed in this
study manipulated the external wall faade at efficient
inclination angles and changed window distribution
for the optimization of daylighting of a south oriented
patient room under the desert clear-sky of Cairo,
Egypt. Grasshopper, a parametric modelling tool was
used to automate the daylighting simulation process.
Daylight analysis was performed by the use of Diva-
for-Rhino plugin via integration with Radiance and
Daysim software. A genetic algorithm was used to
generate and evaluate the performance of the most
successful solutions by automatically adjusting
different variables through control nodes until near
optimum solutions were reached.
Results demonstrated that parametric workflows and
optimization can be effectively used to generate
unconventional patient room faade designs that
provide superior daylighting performance. The
approach proposed in this study produced a wide
range of unconventional faade designs that achieved
100% daylit area on both the bed area and the room
area, with 0% of partially-daylit and over lit areas.
INTRODUCTION
The Hospital patients' rooms are crucial to patient
observation and treatment. Daylighting and external
view can contribute significantly to patients healing,
and help reduce pain and length of stay in hospitals
(FGI, 2010). For optimal daylighting performance, the
external facade should be designed to maximize
daylighting to assist in patient's health care and
comfort.
The positive daylighting contribution to healthcare
was confirmed in several publications (Walch et al.,
2005). Natural light could help reduce stress and
fatigue while increasing effectiveness in delivering
care, patient safety and overall healthcare quality
(Ulrich, 1991 and Ulrich et al., 2004). Pechacek, et al.
(2008) attempted to link environmental cues, such as
lighting, with human performance and health in
healthcare settings. Daylight Autonomy (DA) was
used to simulate the probabilistic and temporal
- 1383 -
potential of daylight for human health needs. The
variability of key architectural decisions in hospital
room design, orientation, window size, and glazing
material were studied for their impact on achieving
threshold values for lighting. The results demonstrated
that modest amounts of glazing could provide a high
degree of circadian stimulus in certain orientations. In
another study in Malaysian hospitals, a review was
conducted on the effect of daylighting on healing the
patients, besides reducing the artificial lighting energy
consumption. The research studied the hospitals'
daylight and artificial lighting and their relation to
other environmental aspects. Methodology included
literature review, desktop analysis and pilot studies of
hospital buildings (Aripin, 2007).
In more related research work, the influence of room
shape on daylighting performance in hospital patient
rooms was investigated. A patient room layout having
an outboard bathroom proved to be most successful in
regards to provision of daylighting in the south
orientation. It provided the largest range of acceptable
window opening sizes and shapes. It offered the
possibility of having large size windows that offer
better external view. However, the patient room
layout with an inboard bathroom results in much
smaller range of acceptable window configurations
(Sherif et al., 2014). In another publication,
optimization of window opening in a hospital patient
room was conducted in a research that aimed at
providing daylighting, external view, while
minimizing the energy consumption. An optimization
methodology was demonstrated through parametric
computer simulations to determine the optimum
window design in the form of window width, sill and
lintel heights and shading device depth (Shikder et al.,
2010).
On other hand, digital design has evolved to
incorporate unique formal content and include
performance based modelling that is reflected on
architectural design considerations. This has led to the
development of complex architectural concepts using
adaptive systems and optimization of forms all
dependent on mathematics (Okabe et al., 2009).
Parametric modeling tools has the power to
manipulate geometry and generate many variations by
changing its parameters which can be done in manual
or automatic way, not only logical operations or other
modeling functions are the limit of parametric
manipulation, the reason is parametric workflow can
Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.

support various geometrical instances in a single
model not like other liner 3D modeling tools (Salim &
Burry, 2010). Parametric design tools can be
categorized into two groups; the first group called
associative-geometry, which characterized by
mathematical associations between the 3D elements
such as points, curves and surfaces. Grasshopper for
Rhino and Generative Components by Bentley are the
tools, which belong to this group. The second group is
focused on Building Information Modeling. This
group is characterized by parametric relationships
between components of a building design that can
across multiple disciplines without too much effort
(Drogemuller et al., 2004). The parametric modelling
tools have graphical and symbolic interface to process
and manage the parametric geometries. However
symbolic diagrams are more often used to define the
logic and constrains. Grasshopper has this feature
presented in various components, parameters,
constraints, and associations, which can assist
designers to form the logic that can be applied on
different projects. Moreover, the key benefit of this
parametric approach is valorized at the early design
stages (Salim & Burry, 2010).
In a study related to this paper, parametric workflow
was proposed to help architects find the optimum
daylighting solutions for different types of buildings.
A workflow, or algorithm, was developed to search for
the optimal dimensions for window openings to ensure
maximum daylighting provision inside a space. This
parametric procedure produced precise and relatively
efficient solutions, automatically without the need to
export or import the 3D modelling information from
different software programs (Wagdy, 2013). Another
parametric optimization workflow was created to test
the external and internal reflectors and ceiling
geometry for a deep side lit space. This workflow used
validated daylight simulation along with genetic
optimization algorithm to find near optimum
configurations for the reflectors that archive adequate
daylighting in Cairo (Wagdy & Shalaby, 2013).
The above review demonstrated that a limited number
of publications were concerned with the optimization
of the geometry patient room external facades for the
improvement of daylighting performance in the desert.
It also provided evidence to the potential of utilizing
parametric workflows and optimization in reaching
faade solutions that can improve daylighting
performance.
OBJECTIVES
This paper aims at exploring the potential of utilizing
parametric workflows and optimization for the
improvement of patient room design. It concentrated
on the optimization of patient room external faade
geometry for control of solar penetration, and thus
improving daylighting performance. This was
achieved through the manipulation of the external
faade wall at efficient inclination angles and change
of window distribution. The approach suggested in
this paper could pave the way for reaching more
sustainable hospital designs that suit the special
conditions of desert locations.
METHODOLOGY
The methodology adopted in this paper aimed to show
the effectiveness of utilizing the parametric approach
in generating different design alternatives, evaluating
and presenting their performance. This approach could
lead the way to the generation of new unconventional
design ideas.
Parametric Workflow
The workflow developed in this study started with the
generation of a 3D model of the patient room using a
three dimensional parametric modeler tool named
Grasshopper, which is a plug-in for Rhinoceros. After
the 3D model was created, materials and other
simulation parameters were assigned using another
plugin called DIVA, which is interfaced with
Radiance and Daysim (Rutten, 2014; Solemma, 2014).
Galapagos, which is a genetic optimization algorithm,
was used to automate and optimize the daylighting
simulations in Grasshopper (Rutten, 2013). While the
simulations were running, a live connection with MS
Excel was established to document the simulation
results and the corresponding model parameter. This
allowed for a case-by-case examination of results. The
following sections provide a detailed overview of the
parametric methodologies that were essential to
optimize the patient room faade design.
Room and facade parameters
A hypothetical patient room design with a nested
bathroom configuration was selected for investigation.
A parametric faade was created for the tested room
by dividing the outside wall into 4 equidistant vertical
segments as shown in Figure 1.
Figure 1 Hypothetical patient room plan
The range of displacement for the main vertical
segments was from (0 to 0.7 m). Each wall segment
could be divided into 2 attached panels, if a small
window fragment was required by the optimization
logic. Three to four horizontal divisions were added to
each wall segment. This allowed for the generation of

- 1384 -
Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.

a variety of faade design alternatives as shown in sky. No external obstruction was assumed to allow for
Figure 2. all possible sun penetration angles.
Assumed room parameters and material are illustrated
in Table 1. Different parametric variables of external
wall inclination angle, window sizes and positions
were controlled by the optimization algorithm. These
parameters were modelled with their actual
thicknesses to reduce the margin of error.
Table 1 Simulation Parameters
Patients Room Parameters
Materials Parameters
Ceiling 80% White Paint
Figure 2 Examples of the variety of the generated Medium Off-
Walls 50%
faade alternatives before optimization Reflectivity White Paint
Ratios Floor 20% Wooden Floor
The division process was connected to window
Medium Off-
generation algorithm, where the logic was defined to Shading 50%
White Paint
ensure that in all cases the first row of panels that close Visual
to the floor and would be unglazed. The shading Window 80% Double Clear
Transmittance
system that is used to block direct sun rays was a Modelling parameters
horizontal sun breaker positioned at a 45 angle
Area 11.84 - 14.43 m2
located on the top edges of the each horizontal panel,
Horizontal Wall Divisions 3 to 4
with two vertical sun breakers on both sides of the wall
edges as shown in Figure 3. Vertical Wall Divisions 1 to 2 per wall segment
For conceptualization, the parametric workflow was Wall Segment 4 segments
set up to define the control of window configuration Wall Segment
0.7 m
and wall shape variations. As illustrated in Figure 3, transformation limits
every part of the 3D model was generated based on Minimum number of solid First row of horizontal
Divisions divisions (4 to8)
specific geometrical relations and constraints,
Outside shading system 45 cut-off angle
therefore the design variations were accurately
generated. Daylight evaluation criteria
Testing was conducted for year-round performance
using the "Dynamic Daylight Performance Metrics
(DDPM)". The DIVA plugin was used to perform the
daylight analysis via integration with Radiance and
DAYSIM (Reinhart & Wienold, 2011). DIVA (which
stands for Design Iterate Validate Adapt) is an
environmental analysis plugin for the Rhinoceros 3D
Nurbs modeling program (McNeal, 2015). The
occupied time for daylighting simulation was from
8AM to 6PM. Simulations were carried out at 300 Lx
on the bed plane, where measurement was calculated
for three points spaced at a grid of 0.7 m intervals on
the bed surface and the rest of analysis points covered
the rest of the room.
Three Daylight Availability evaluation levels were
used: daylit, partially daylit and over lit areas.
The daylit areas are those areas that received
sufficient daylight at least half of the year-round
occupied time. The partially daylit areas are those
areas that did not receive sufficient daylight at least
half of the year-round occupied time. The over lit
areas are those areas that received an oversupply of
Figure 3 Parametric modelling processes daylight, where 10 times the target Illuminance was
The room was assumed to be located on the second reached for at least 5% of the year-round occupied
floor level of a hospital building in the outskirts of time (Reinhart & Wienold, 2011).
Cairo, Egypt, which enjoys a year-round desert clear- The acceptance criterion adopted in this paper
assumed that the cases where the daylit area reached

- 1385 -
Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.

50% of the tested space were considered having
acceptable performance. Cases having 100% daylit
area over the bed coupled with more than 90% daylit
area of the Patient room area were considered
successful, since these ensure sufficient daylighting
performance at the patient bed where medical care is
performed.
RESULTS AND DISCUSSION
Use of the optimization method proposed in this study
succeeded in providing a wide range of design
variations. This allowed for showing the effectiveness
of using the proposed parametric approach in
identifying a large number of unconventional designs
that maintain maximum daylighting performance.
This is demonstrated in Figure 4.
The optimization process was finished after 262
simulation runs. As shown in Figure 4, simulation
results were ordered from the worst performance to
the optimized ones that achieved 100% daylit area on
both the bed area and the room area, with 0% of
partially-daylit and overlit areas. The optimum
solutions had an almost similar daylighting
performance, even though they were completely
different in regards to the faade wall configuration.
Samples of the most unique of these configurations are
illustrated in Figure 5.
The genetic algorithm adopted in this study enhanced
daylighting performance during the optimization
process gradually through 29 Generations, by which
the objective was set to maximize daylit area
percentage and to minimize the over-lit and partially-
daylit area percentages. In the last generation, by the
end of the optimization process, 42 optimum solutions
were defined. These all met the targeted criteria.
CONCLUSION
This paper demonstrates that parametric workflows
and optimization can be effectively used to generate
unconventional patient room faade designs that
provide superior daylighting performance. The
approach developed in this study manipulated the
external wall faade at efficient inclination angles and
changed window distribution for the optimization of
daylighting.
A large number of solutions were developed, where
100% daylit area on both the bed area and the room
area, with 0% of partially-daylit and overlit areas were
achieved. These results were not attainable with the
use of traditional methods of simulation by trial and
error. Experiments showed the effect using automatic
optimization techniques in manipulating wall
configuration in order to achieve 100% daylit spaces.
The approach proposed in this study produced a wide
range of unconventional faade designs. Apart from
conventional 2D plane facades, new results were
developed where formal design aspects were
integrated with the requirements of daylighting
performance. Using this approach, a new set of
unexpected design solutions that comply with
daylighting requirements was generated. Thus, the
proposed methodology provided another design
perspective that could be used in the early design
stages, where decisions leverage both form aspiration
and performance limitation.
The proposed approach is currently limited to
daylighting aspects. It does not account for other
important performance factors such as acoustic,
thermal and external view exposure. Addition of these
aspects to the optimization process will help narrow
down the alternatives to the ones that comprehensively
achieve a balanced performance.
ACKNOWLEDGEMENT
This publication is based on research work supported
by a research grant from The American University in
Cairo.
Optimum solutions

Figure 4 Optimization results according to Daylight Availability metric

- 1386 -
Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.

Figure 5 Facade configuration based on the optimum results

- 1387 -
 Proceedings of BS2015:
14th Conference of International Building Performance Simulation Association, Hyderabad, India, Dec. 7-9, 2015.

REFERENCES
Solemma, L. (2014). DIVA-for-Rhino software
Aripin S. (2007). Healing Architecture: Daylight in
version 3.0. Retrieved 5 September 2014, from
hospital design. Proceedings of the Sustainable
http://www.diva4rhino.com
Building Conference. South East Asia. Malaysia.
Ulrich, R. (1991). Effects of healthcare interior design
Drogemuller, R. M., Crawford, J., & Egan, S. (2004).
on wellness: Theory and recent scientific
Linking early design decisions across multiple
research. Journal of Healthcare Interior Design,
disciplines. Paper presented at the Product and
vol. 3, 97-109.
Process Modelling in the Building and
Construction Industry (ECPPM 2004), Istanbul, Ulrich, R. and Zimring, C. (2004). The role of the
Turkey. physical environment in the hospital of the 21st
century: a once-in-a-lifetime opportunity. The
FGI (2010). Guidelines for Design and Construction
Center for Health Design SM.
of Healthcare Facilities. USA: Facility Guidelines
Institute. Wagdy, A. (2013). New parametric workflow based
on validated day-lighting simulation. Proceedings
McNeel, R. (2015, March 2). Rhinoceros Version 5.0
of the Building Simulation Cairo 2013, Cairo,
Service Release 10. Retrieved from
Egypt.
http://www.rhino3d.com/
Wagdy, A. & Shalaby, M. (2013). Optimizing the
Okabe A, Boots B, Sugihara K, Chiu SN. Special
external and internal reflectors and ceiling
tessellations: concepts and applications of
geometry for a deep side lit space: Using
Voronoi diagrams. Chichester: Wiley; 2009
validated daylight simulation with genetic
Pechacek, C., Andersen, M. and Lockley, S. (2008). optimization algorithm in Cairo. Proceedings of
Combining annual daylight simulation with the Sustainable Building Conference SB13
photobiology data to assess the relative circadian CAIRO, Cairo, Egypt.
efficacy of interior spaces. Proceedings of the 5 th Walch, JM., Rabin, BS., Day, R., Williams
National Conference of IBPSA. eSim 2008. JN., Choi, K., and Kang, JD. (2005). The effect
Canada. of sunlight on postoperative analgesic
Reinhart, C. F., & Wienold, J. (2011). The daylighting medication use: A prospective study of patients
dashboarda simulation-based design analysis for undergoing spinal surgery. Psychosomatic
daylit spaces. Building and Environment, 46(2), Medicine, 67 (1), 165-163.
386-396.
Rutten, D. (2013). Galapagos: on the logic and
limitations of generic solvers. Architectural
Design, 83(2), 132-135.
Rutten, D. (2014). Grasshopper-Algorithmic
modeling for Rhino software version 0.9077.
Retrieved 5 September 2014, from
http://www.grasshopper3d.com
Salim, F. D., & Burry, J. (2010). Software openness:
evaluating parameters of parametric modeling
tools to support creativity and multidisciplinary
design integration Computational Science and
Its ApplicationsICCSA 2010 (pp. 483-497):
Springer.
Sherif, A., Sabry, H., and Wagdy W. (2014). Hospital
Patient Room Design for Desert Climates: Effect
of Room Shape on Window Design for
Daylighting, Proceedings of the Second Saudi
Forum for Planning and Design of Hospitals.
Saudi Arabia.
Shikder, S., Mourshed, M. and Price, AD. (2010).
Optimisation of a daylight-window: Hospital
patient room as a test case. Proceedings of the
International Conference on Computing in Civil
and Building Engineering. Nottingham. UK.

- 1388 -

View publication stats