Font Size: 

Introduction

“Garbage in, garbage out” is a familiar expression which means how well a decision is made based on accurate data and analysis will determine how well will be the output of the product (i.e. design). In architectural design process, Pena and Parshall (2012) differentiate programming and design succinctly as problem seeking and problem solving; In order to solve the problem well, the problem needs to be well defined. Because of the nature of problem seeking, Pena and Parshall, as well as other programing publications, offers fairly generic procedural guidance without restricting any order of the steps and methods to use. The good thing about it is that allows programmers to think analytically but also creatively, without restriction. The concern though, with increasing complexity of building function, form, economy, and time, programmers would face difficulties in quickly reaching down to the essence of the problem. This study proposes a formalized computational method of analyzing one of the building functions in programming: space adjacency by integrating various information and techniques from different fields; and then evaluates whether the method could be applicable in practice. The case study used is one from healthcare building type as it is arguably one of the most complicated building types (Kumar, 2011).

The development of healthcare facilities in the United States has been impacted by the healthcare reform known as the Affordable Care Act, which expects these facilities to lower their cost of service and to ensure better quality of care (medicaid.gov, 2015). In order to meet these objectives, people who are experts in the field of healthcare systems often implement patient flow analyses by constructing patient, personnel, material and logistic flows, as well as information process flows.  In the field of architecture, space adjacency analysis is implemented to improve flow of movement from one space to another, related to the function of the space. In other words, the rationale of space adjacency analysis is to improve flow between spaces (White, 1986). The output of a space adjacency analysis, called an adjacency/relationship matrix, is then used as an input for space planning layout, which also plays an important role in the quality of services on the operational as well as the human factor level (Arnolds, et. al., 2012). While both fields from healthcare system and architecture view process flow as either a measurement or an objective of their analysis, there is a lack of systematic approach to integrate process flow as a base of measurement and development of space adjacency.

As part of building programming, any computational approach to space adjacency analysis should assist programmers’ ways of thinking (Pena and Parshall, 2012), specifically in objective thinking, comprehensive views, and feedforward - evaluation activity within a heuristic nature. The current process for analyzing adjacency requirements often relies on experience and expert opinions, which are somewhat subjective; it has developed to a more systematic and less intuitive approach, for example using techniques such as travel charting and the number of journeys made, although this data might not be readily available in designing a new facility (Foulds, 1983).

An emerging body of literature on computational space planning layout shows a type of approach that assists in objective thinking, but limits comprehensive views such as correlation between configuration layout and social or organizational implications. Some of the computational techniques that implement layout optimization restrict programmers’ feedforward - evaluation process to "...toughtlessly trying out random arrangements of spaces..." (Nourian, et.al., 2013). The feedforward - evaluation process mostly starts by determining an initial idea of the relationships among spaces and then evaluating it. The relationships can then be adjusted and evaluated based on other ideas or different factors and then evaluated again until the programmers reach condition(s) that allow the development of creative design solutions (Pena & Parshall, 2012). Foulds (1983) suggested that computational approach in graph theory is a promising approach because it can deliver an easy-to-understand benchmark for the upper bound values of feasible space relationship arrangements.

Justified Plan Graph (JPG) is a graph form that has been applied in architecture as part of Space Syntax theory. A theory that is interested in the correlation of space configuration with social or organizational aspects (Ostwald, 2011). The basic method is to transform spaces and their access to other spaces in a floor plan layout into topological graphs as nodes and lines; keep clear from examining geometrical aspects of a floor plan such as length, width, and distance.  These graphs then analyze in terms of connectivity and integration of spaces, as well as other measurements that have been developed as ways to explain spatial, social, or organizational characteristics. This method of somewhat ‘stripping down’ the complexity of a floor plan into configuration of spaces, seems to resemble similar structure for space adjacency analysis, where some geometrical aspects have not yet been developed in this early stage of architectural process. Although Space Syntax theory has been frequently used  for the analysis of already drawn layouts and built environments in correlation with social and organizational effects (Choudhary, et. al., 2010), little attempt has been made to implement Space Syntax as a part of space adjacency layout and as a basis for developing facility layout designs.

Methodological Procedures and Results

This paper will present an initial attempt to integrate into the space adjacency analysis process (a) the information and techniques of activity flow diagrams, which are often found in healthcare system operations; and (b) Space Syntax JPG analysis, which is often used in building configuration analysis. The two primary research questions are:

1. How do we formalize a process of integrating information and techniques from various fields into a computational space adjacency analysis system?

2. How does the output of this assistive system compare with real, designed layouts?

A case study of a cancer clinic that is currently at the construction phase is used. As the support data for addressing the first research question, information from the building programming process of the cancer project was collected. The building programming goal for the cancer clinic project is to make the nurse station area the most accessible area from the other spaces in the clinic. For addressing the second research question, the floor plan from the actual project has been collected as support data.

To address the first research question, a UML data model is developed to describe the methodology of the integration of information and techniques from activity flow and space syntax  into the computational space adjacency analysis system. Activity flows of stakeholders (patients, physicians, and nurses) are developed with a common commercial-off-the-shelf (COTS) software for healthcare system operation. Spaces are identified and relationships among spaces are assigned in accordance to the activities in the flow. These spaces and relationships are exported into an architectural visual programming software and visualized as a graph. Space Syntax JPG analysis is developed in this architectural software to analyze the graph and provide the ability to adjust the space relationships in real-time until acceptable conditions that meet the building programming’s objective are reached.

The finalized data model of the formalized process is illustrated in the results section along with the successes and challenges encountered in the development of the system. Although all the integration of information and techniques could be done on one visual programming platform, the experiment illustrates a collaboration scenario with a healthcare system field by showing an integration process of software platforms from the two fields. An adjacency matrix is the output of the analysis, and an initial attempt is also made to illustrate the implementation of the matrix into a bubble diagram.

To address the second research question: Once several acceptable conditions are reached in the form of space syntax analysis, the space syntax analysis from these conditions is compared with the space syntax analysis from the floorplan of the real project. Because the floorplan of the real project has more detailed and completed space programs, a strategy is developed to assure the validity of the comparison measurement. Any similarities and discrepancies are analyzed and explained.

Discussion

Healthcare facilities are known to be complicated building types, not least because  they must provide the complex functions of healthcare services. Many specialized consultants from different fields play an important role in the planning and design of these facilities. This study proposes an initial concept for integrating information and techniques from the fields of healthcare systems and of architecture into the building programming process. Although this study focuses on the development of space adjacency - space layout, the process could be worked backwards, in which the manipulation of space adjacency - space layout could potentially be used to adjust the operational flow of the facilities.

The procedure of current building programming processes involves different orders of steps (goals, facts, concept, and needs) depending of the size and stage of the project, the availability of the information, stakeholders, etc. This study proposes a formalized method of space adjacency analysis along with the computational tools to support it without discarding programmer's’ creative and analytical way of thinking.

 

spatial adjacency; process flow; space syntax; graph theory; building programming