Research on the Suitability of Building Integrated Agriculture—Taking Indoor Living Walls as an Example

Abstract

As urbanization accelerates and the availability of arable land declines sharply, building-integrated agriculture (BIA) has emerged as a crucial strategy for enhancing urban food security and it also promotes the establishment of sustainable urban food production systems. This study focuses on indoor living walls (ILWs) and employs the analytic hierarchy process (AHP) and the entropy weighting method to construct a comprehensive suitability evaluation model. The model evaluates different spatial layouts across five key dimensions: indoor microenvironment, physiology of vegetable, morphology of vegetable, yield of vegetable and quality of vegetable. The results reveal that among the experimental groups, R2 was classified as suitable, with an average group score of 2.29. The remaining groups were classified as moderately suitable, with descending average scores of 1.64 for R3, 1.43 for R4, and 1.16 for R1. Based on the climatic characteristics of Hainan Province, the optimal configuration is recommended to include a north-facing room, a west-wall planting layout, and a “partial human–vegetable separation” spatial strategy, with an installation height exceeding 1.3 m.

1. Introduction

With the rapid urbanization, the human living environment faces unprecedented challenges. High-intensity urban expansion continues to encroach on cultivated land. Urban development in populous countries such as China has consumed 33,080 km² of agricultural land, further exacerbating the food security crisis [1,2]. In response to this challenge, building-integrated agriculture (BIA) has emerged as an innovative solution. At its core, BIA integrates food production into the building envelope and interior spaces through interdisciplinary system design and technological innovation, promoting synergistic coexistence between architectural functions and agricultural production [3]. BIA establishes a new paradigm of “building–agriculture” interaction. On one hand, buildings offer physical space and environmental regulation to support agricultural processes [4]. On the other hand, agricultural systems enhance building performance by providing ecosystem services such as transpiration-based humidification [5], shading and cooling [6,7], carbon sequestration and oxygen generation [8].

BIA mainly includes two types: rooftop farm (RF) and indoor farm (IF) [4,9]. Among them, indoor farms (IF), characterized by vertical planting, such as plant factories, have developed particularly rapidly in the agricultural field [10] Plant factories represented by Aero Farms of the United States [11], Plant Lab of the Netherlands [12] and Sananbio of China [13] generally adopt soilless cultivation technology, effectively achieving the goals of low resource consumption and high crop output. In the field of architecture, the headquarters building of the Pasona company fully demonstrates the deep integration of architecture and agriculture. This project has maturely applied the concept of indoor farms, systematically integrating over 200 types of crops into the building’s exterior envelope and internal office Spaces [14].

In the field of architecture, indoor agricultural planting in buildings mainly exists in the form of indoor living walls (ILWs). Research related to ILWs is limited, and the existing achievements mostly focus on aspects such as technical implementation, social benefits, and energy benefits [15,16,17,18,19,20]. Current research not only lacks a refined consideration of the installation location of ILWs, but also neglects the complex dynamic interaction mechanism between ILWs and the physical environment of the building in which it is embedded as a whole. Even within the same building, environmental factors such as temperature, humidity, illumination intensity, and CO₂ concentration can vary significantly across different spatial zones [21]. Neglecting these microenvironmental variances may result in plants experiencing long-term suboptimal or even stressed conditions, which can diminish expected ecological and aesthetic outcomes, elevate maintenance costs, and reduce system efficiency—ultimately leading to abandonment. This is not conducive to the development of ILWs. Therefore, determining the optimal spatial layout of ILWs is a key task for promoting urban food safety and the widespread application of BIA. This study constructs a multi-criteria evaluation model to quantitatively assess the suitability of different spatial layouts. Utilizing the evaluation model during the planning and design phase of BIA to provide technical support for planning, site selection, and operations will promote the application and development of BIA.

This study conducted a suitability assessment of ILWs using a building located on the campus of Hainan University in Haikou, Hainan Province, as a case study. Based on the building’s indoor environmental characteristics and plant growth conditions, a suitability evaluation model was constructed using the AHP combined with the entropy weight method. By determining indicator weights and calculating comprehensive evaluation scores, the study assessed the environmental suitability of ILWs installed in indoor spaces with varying spatial layouts. The findings provide a scientific basis for informed site selection and system design in BIA projects. This enables the proactive identification of optimal environments that enhance plant survival rates, growth performance, and ecological service functions of ILWs, thereby improving the system’s overall ecological and economic efficiency, as well as its long-term operational sustainability.

2. Materials and Methods

2.1. Design of Indoor Living Walls

This study selected a building at Hainan University in Haikou, Hainan Province, as the experimental site for lettuce cultivation. The experiment monitored a total of 18 indicators across five aspects: indoor micro-environment B1, physiology of vegetable B2, morphology of vegetable B3, yield of vegetable B4, and quality of vegetable B5. The experiment was designed with four groups of offices on the same floor but with different spatial layouts. Among these, Room1 (R1) and Room2 (R2) are north-facing rooms, while Room3 (R3) and Room4 (R4) are south-facing rooms. The ILWs for R1 and R3 were placed on the east wall, while those for R2 and R4 were placed on the west wall, as shown in Figure 1.

The ILWs in this study were designed with a four-layer vertical planting system, with planting tanks arranged at equal intervals and a vertical spacing of 300 mm, as shown in Figure 2. To investigate the effect of different heights on planting performance, the four-layer planting tanks were divided into two experimental sample groups based on height: the lower layer (ground height 0.4~0.85 m) was Group A, and the upper layer (ground height 1.3~1.75 m) was Group B. The experiment was set up with four independent units, each containing two experimental groups (A and B), totaling eight experimental sample groups. The sample groups were labeled using the format “experimental unit number—group,” with detailed grouping information provided in

Table 1. Experimental sample group settings.

Sample Group	Room Orientation	Set of ILWs	Height
R1-A	North	West	0.4~0.85 m
R1-B	North	West	1.3~1.75 m
R2-A	North	East	0.4~0.85 m
R2-B	North	East	1.3~1.75 m
R3-A	South	East	0.4~0.85 m
R3-B	South	East	1.3~1.75 m
R4-A	South	West	0.4~0.85 m
R4-B	South	West	1.3~1.75 m

.

The planting system consists of planting tanks, mounting brackets, an irrigation system, and artificial lighting, as shown in Figure 2a. The planting tanks are made of PVC square tubes with dimensions of 200 mm × 150 mm. Each tube is uniformly perforated with 11 planting holes, each with a diameter of 80 mm, spaced 200 mm apart. These planting tanks are securely fixed to the wall using T-shaped brackets, with installation details shown in Figure 2b. The irrigation system includes a water storage tank, pump, drip irrigation, and water pipes. The system is equipped with a smart plug, allowing irrigation timing to be controlled by a mobile app. The artificial light uses LED lights with a length of 1.2 m, power of 18 W, and a red-to-blue light ratio of 3:1. Two LED lights are installed in parallel on each layer of the planting wall, with automated control by smart plugs, operating daily from 7:00 p.m. to 7:00 a.m. the next day. The cultivation matrix uses a composite substrate made by mixing turfy soil, perlite, and vermiculite in a 1:1:1 volume ratio.

2.2. Materials

This study selected butter lettuce (Lactuca sativa L. var. capitata) as the experimental crop, with seeds purchased from Shenyang Xiangrui Agricultural Technology Co., Ltd., Shenyang, China. Prior to the experiment, the seeds were pretreated, then sown into 1-inch rockwool seedling blocks using nutrient solution for seedling cultivation. During the seedling stage, the temperature was maintained at 23 ± 2 °C. Lighting was provided by LED plant growth lights of the same specifications as those used in the vertical farming system, with a photoperiod set at 12 h·d⁻¹. On 28 December 2024, the lettuce seedlings were transplanted into each experimental sample group. Each experimental sample group was uniformly planted with 22 lettuce plants. Samples were uniformly harvested on 8 February 2025.

2.3. Measurement Methods

2.3.1. Measurement of Indoor Microenvironment Indicators

Indoor microenvironmental indicators were measured for eight experimental sample groups on 29 December, 20 January, and 7 February, respectively. The measurement points were arranged as shown in Figure 3. The final data were calculated as the arithmetic mean of three measurements. Indoor temperature and humidity were measured using the HOBO UX100-003 temperature and humidity recorder produced by ONSET (Bourne, MA USA) in the United States. Monitoring was conducted from 8:00 to 18:00, with data recorded once per hour. The indoor CO₂ concentration was measured using the Telaire-7001 CO₂ concentration monitor manufactured by GE (Boston, MA, USA) in the United States. The monitoring time was from 8:00 to 18:00, with one measurement recorded per hour. The indoor illumination intensity was measured using a TES1330A digital lux meter manufactured by TES Electrical Electronic Corp in Taipei, China., with monitoring conducted from 8:00 to 18:00, recording once per hour. All measurement data were calculated as the arithmetic mean of three repeated measurements to obtain the final results.

2.3.2. Measurement of Vegetable Indicators

Physiological indicators were measured on 21 January, and vegetable morphology, yield, and quality were measured on 8 February. Vegetable sampling was conducted using random sampling, with three lettuce plants randomly selected from each experimental sample group for indicator measurement. In physiological and biochemical experiments, lettuce leaves were divided into four parts to create test samples, with three replicates for each test sample. The final data were the average of the results from the three replicates. The indicator measurement methods are as follows:

(1)

Morphological indicators of lettuce

Use a ruler with an accuracy of 1 mm to measure plant height and plant breadth. Plant height is measured from the root–stem junction to the growing point. Plant breadth is measured as the maximum width of the aboveground part of the plant. Use an electronic vernier caliper with an accuracy of 0.1 mm to measure stem diameter, with the measurement point located 1 cm below the cotyledon. Use the easy leaf area application developed and applied by the University of California to measure maximum leaf area [22].

(2)

Physiological indicators of lettuce

Spectrophotometry was used to measure chlorophyll content in lettuce leaves and root vitality. The test kits were purchased from Suzhou Grace Biotechnology Co., Ltd., Suzhou, China, and the measuring instrument was a UV-1100 visible light spectrophotometer.

(3)

Yield indicators of lettuce

Vegetable yield was measured using an electronic scale with an accuracy of 0.01 g. Note that the leaves and roots were washed with clean water, dried with absorbent paper, and then separated into the leaves and roots of the lettuce, which were weighed separately to obtain the fresh weight of the above-ground and below-ground parts.

(4)

Quality indicators of lettuce

Spectrophotometry was used to measure vitamin C content, soluble sugar content, and nitrate nitrogen content. The test kits used for the measurements were purchased from Suzhou Grace Biotechnology Co., Ltd. Vitamin C and soluble sugar content were determined using a UV-1100 visible light spectrophotometer, while nitrate nitrogen content was determined using a U-T3C ultraviolet-visible spectrophotometer.

2.4. Selection of Evaluation Methods

Currently, widely used comprehensive evaluation methods include the analytic hierarchy process (AHP) [23], fuzzy mathematics [24], the entropy weight method (EWM) [25], principal component analysis (PCA) [26], the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) [27], and the Delphi method [28], among others. Each method has distinct advantages and limitations depending on the evaluation dimensions and data characteristics. In existing research, the analytic hierarchy process (AHP) has been widely applied in comprehensive evaluations across multiple fields. Researchers have effectively utilized this method to decompose complex problems into layers, making it particularly suitable for constructing indicator systems [29,30,31,32,33]. However, in terms of weight calculation, AHP is prone to bias due to the subjective scoring of experts [34,35,36]. The core of fuzzy mathematics lies in quantifying the degree to which an object belongs to a fuzzy set through the construction of membership functions [37]. This method performs exceptionally well in scenarios where the concept itself is inherently fuzzy, involves continuous processes, or deals with complex systems. However, its effectiveness largely depends on the appropriate selection of membership functions. The entropy weight rule objectively assigns weights based on the dispersion of indicator data, with its core application scenario being the use of data information structure to judge the importance of indicators. PCA can reduce the number of variables and eliminate multicollinearity issues among indicators by extracting principal components through dimensionality reduction, making it suitable for scenarios with numerous indicators and strong correlations [38]. TOPSIS ranks the proximity of each scheme to the ideal solution to determine their relative merits, making it suitable for multi-scheme comparisons [39].

After analyzing the applicability of various methods, this paper ultimately selected the AHP–entropy weighting method model, which combines AHP with the entropy weighting method. On the one hand, this model utilizes AHP’s hierarchical decomposition to evaluate objectives. On the other hand, by using the entropy weighting method to correct objective data, it can effectively balance subjective judgment with the objective distribution characteristics of the data [40,41]. The combination of the two methods ensures the scientific rigor of the evaluation system while maintaining a balance between objective and subjective factors, thereby enabling a more precise assessment of the environmental suitability of ILWs.

2.5. Establishment of Evaluation Model

This study adopts an approach focused on the suitability evaluation of ILWs, comprehensively considering indoor environmental factors and the influencing factors throughout the entire plant growth process, ultimately identifying 18 evaluation indicators. The overall evaluation objective A is divided into five criterion-level categories: indoor microenvironment B1, physiology of vegetable B2, morphology of vegetable B3, yield of vegetable B4, and quality of vegetable B5. These five criteria are further subdivided into 18 specific indicators: temperature C1, CO₂ concentration C2, humidity C3, illumination intensity C4, root vitality C5, total chlorophyll C6, chlorophyll a C7, chlorophyll b C8, stem diameter C9, number of blades C10, maximum leaf area C11, plant height C12, plant breadth C13, shoot fresh weight C14, root fresh weight C15, vitamin C content C16, soluble sugar content C17, and nitrate nitrogen content C18. The complete hierarchical structure is illustrated in Figure 4.

2.6. Determination of Evaluation Indicator Weights

2.6.1. Determination of AHP Weights

(1)

Construct comparison matrix

Using the Saaty 1–9 scale method, pairwise comparisons are conducted across all levels of indicators to construct a comparison matrix. The scale values, as shown in

Table 2. Meaning of Saaty 1–9 scale values.

Value	Meaning
1	The two elements are equally important when compared
3	The former is slightly more important than the latter
5	The former is significantly more important than the latter.
7	The former is extremely important compared to the latter
9	Indicates that the former element is more important than the latter
2, 4, 6, 8	The median value of the above adjacent judgments
The reciprocal of 1 to 9	The importance of comparing the order of the corresponding two factors

, represent relative importance: a score of 1 indicates equal importance between two elements; 3 denotes that the former is slightly more important than the latter; 5 signifies a strong degree of importance; 7 reflects a very strong level of importance; and 9 indicates that the former is extremely more important than the latter. The intermediate values 2, 4, 6, and 8 are used to express compromises between these judgments.

(2)

Weight calculation

Assume that the comparison matrix $A=\left[\begin{array}{ccc}{a}_{11}& \dots & a_{n1}\\ ⋮& \ddots & ⋮\\ a_{n1}& \dots & a_{nn}\end{array}\right]$, and calculate the weights using the square root method, as Equations (1) and (2).

$${\varpi }_i=\sqrt[n]{{\prod }_{j=1}^n{a}_{ij}}.$$

(1)

$${\omega }_i={\displaystyle \frac{{\varpi }_i}{{\sum }_{j=1}^n{\varpi }_j}}.$$

(2)

(3)

Consistency check

After calculating the weights of each indicator in the comparison matrix, the consistency check is performed. The maximum eigenvalue ${\lambda }_{max}$ is calculated according to Equation (3), and after obtaining ${\lambda }_{max}$, the consistence index (CI) value is calculated according to Equation (4). The random index (RI) value is determined based on the order of the matrix, as shown in

Table 3. Average random index.

n	1	2	3	4	5	6	7	8	9	10	11	12	13
RI	0	0	0.58	0.9	1.12	1.24	1.32	1.41	1.45	1.51	1.49	1.54	1.56

n refers to the order of the judgment matrix, which is the number of factors involved in the comparison matrix. RI denotes the Random Consistency Index. They are used to perform consistency tests on the judgment matrix, ensuring the logical reliability of decision-makers’ judgments.

. Finally, the value of consistence ratio (CR) is calculated according to Equation (5). When CR < 0.1, the comparison matrix passes the consistency check and the weight distribution is reasonable.

$${\lambda }_{max}={\displaystyle \frac{1}{n}}{\sum }_{i=1}^n{\displaystyle \frac{{\left(AW\right)}_i}{W_{ai}}}.$$

(3)

$$CI={\displaystyle \frac{{\lambda }_{max}-n}{n-1}}.$$

(4)

$$CR={\displaystyle \frac{CI}{RI}}.$$

(5)

2.6.2. Determination of Entropy Weights

(1)

Standardization of indicators

Assuming that the data matrix is $X=\left[\begin{array}{ccc}{x}_{11}& \dots & x_{m1}\\ ⋮& \ddots & ⋮\\ x_{n1}& \dots & x_{nm}\end{array}\right]$, standardization is performed using Formulas (6) and (7) based on the data characteristics. Positive indicators are those for which larger values contribute more significantly to the evaluation target, such as illumination intensity, chlorophyll content, and plant height. Negative indicators are those for which smaller values contribute more significantly to the evaluation target, such as nirate nitrogen content.

Positive indicator standardization formula

$$X_{ij}={\displaystyle \frac{x_{ij}-min\left\{X_i\right\}}{max\left\{X_i\right\}-min\left\{X_i\right\}}}.$$

(6)

Negative indicator standardization formula

$$X_{ij}={\displaystyle \frac{{max\left\{X_i\right\}-x}_{ij}}{max\left\{X_i\right\}-min\left\{X_i\right\}}}.$$

(7)

(2)

Calculate entropy weights to determine weights

Then, substitute the standardized results into Equation (8) to calculate the entropy weight, and finally calculate the entropy weight weighting according to Equation (9).

$$e_j=-{\displaystyle \frac{1}{ln\left(n\right)}}{\sum }_{i=1}^n{\displaystyle \frac{x_{ij}}{\sum _{i=1}^m{x}_{ij}}}\mathrm{l}\mathrm{n}\left({\displaystyle \frac{x_{ij}}{\sum _{i=1}^m{x}_{ij}}}\right).$$

(8)

$${\omega }_j={\displaystyle \frac{1-e_{\mathrm{j}}}{\sum _{i=1}^n{e}_{\mathrm{j}}}}.$$

(9)

2.6.3. Determination of AHP–Entropy Weight Combination Weights

The weights of each indicator, ${\omega }_i$ and ${\omega }_j$ are calculated using the AHP method and entropy weight method described above. Finally, the composite weight W is calculated according to Equation (10).

$$W={\displaystyle \frac{{\omega }_i{\omega }_j}{{\sum }_{i=1}^n{\omega }_i{\omega }_j}}.$$

(10)

2.7. Calculation of Suitability Scores

A three-tier classification system was employed to assess the suitability of ILWs. According to relevant standards and guidelines, evaluation criteria were categorized into three levels: Suitable I, Moderately Suitable II, and Unsuitable III. Measured data were then compared against the threshold ranges for each level. A score of 1 was assigned if the data met the Suitable I standard, 2 for Moderately Suitable II, and 3 for Unsuitable III. Based on these values, the final suitability scores for each experimental group were calculated using Equation (11). The resulting ILW suitability scores ranged from 0 to 3, with values between 0 and 1 classified as Unsuitable III, 1 to 2 as Moderately Suitable II, and 2 to 3 as Suitable I.

$$G={\sum }_{i=1}^n{w}_i\times g_i.$$

(11)

$G$—The suitability scores of ILWs

$w_i$—Combination weights

$g_i$—Assigned value

3. Results

3.1. Standard of Suitability Evaluation

Plant growth is dependent on maintaining environmental conditions within a specific optimal range; deviations beyond this threshold can hinder growth or even result in plant mortality. Drawing from plant physiology books and the relevant literature, the environmental requirements for lettuce cultivation were classified into three suitability levels—Suitable I, Moderately Suitable II, and Unsuitable III. It is based on four key environmental variables: temperature, CO₂ concentration, humidity, and illumination intensity. For indicators lacking established standards, such as morphology of vegetable, physiology of vegetable, and quality of vegetable, the Jenks method was employed to define grading criteria. The Jenks method is a one-dimensional data clustering technique that determines natural groupings by minimizing intra-class variance while maximizing inter-class variance [42]. In accordance with European food safety standards, nitrate content in vegetables was divided into three regulatory levels. Similarly, vitamin C content and soluble sugar levels were classified into three categories based on relevant nutritional guidelines. The final evaluation criteria for all indicators are presented in

Table 4. Indicator classification standards. All indicators are classified into three grades based on the relevant literature and certain methods, namely Suitable I, Moderately Suitable II, and Unsuitable III.

Indicator	Assignment Range
	Suitable I	Moderately Suitable II	Unsuitable III
Temperature C1 [43]	22~25 °C	18~22 °C	<18 or >25 °C
CO2 concentration content C2 [44]	≤1500 ppm	≤1000 ppm	≤350 ppm
Humidity C3 [44,45,46,47]	70~80%	40~70 or 80~85%	<40 or >85%
Illumination intensity C4 [48,49,50]	16~20 klx	<16 or >20 klx	<1.5 or >25 klx
Root vitality C5	>82.508 μg/h/g FW	58.015~82.508 μg/h/g FW	≤58.015 μg/h/g FW
Total chlorophyll C6	>1.061 mg/g FW	0.842~1.061 mg/g FW	≤0.842 mg/g FW
Chlorophyll a C7	>0.601 mg/g FW	0.482~0.601 mg/g FW	≤0.482 mg/g FW
Chlorophyll b C8	>0.492 mg/g FW	0.396~0.492 mg/g FW	≤0.396 mg/g FW
Stem diameter C9	>8.7 mm	5.3~8.7 mm	≤5.3 mm
Number of blades C10	>9	>8	≤7
Max leaf area C11	>153.77 cm2	84.84~153.77 cm2	≤84.84 cm2
Plant height C12	>20.4 cm	15.8~20.4 cm	≤15.8 cm
Plant breadth C113	>30.1 cm	19.7~30.1 cm	≤19.7 cm
Shoot fresh weight C14	>33.59 g	17.86~33.59 g	≤17.86 g
Root fresh weight C15	>2.4 g	1.31–2.4 g	<1.31 g
Vitamin C content C16 [51,52,53]	>0.67 mg/g	0.61~0.67 mg/g	<0.61 mg/g
Soluble sugar content C17 [53]	>11.8 mg/g	8.1~11.8 mg/g	<8.1 mg/g
Nitrate nitrogen content C18 [54,55]	<3000 mg/kg	3000~3500 mg/kg	>3500 mg/kg
Assignment 1	3	2	1

¹ The three grades are respectively assigned values of III, II, and I. Compare the measured data with the data range of each grade to determine its grade and assign the corresponding score.

.

3.2. Calculation of Evaluation Indicator Weights

3.2.1. AHP Weights

(1)

Objective layer A

The objective layer A includes five indicators: indoor micro-environment B1, physiology of vegetable B2, morphology of vegetable B3, yield of vegetable B4, and quality of vegetable B5. All five indicators are essential for assessing the overall suitability of ILWs, and a deficiency in any single dimension can significantly compromise the final vegetable output. These indicators are considered equally important in the evaluation process. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 5, CR = 0 < 0.1. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final A-Bi weights are shown in

Table 5. Comparison matrix of objective layer.

A	B1	B2	B3	B4	B5	$\mathit{\omega }$
B1	1	1	1	1	1	0.2
B2	1	1	1	1	1	0.2
B3	1	1	1	1	1	0.2
B4	1	1	1	1	1	0.2
B5	1	1	1	1	1	0.2
${\lambda }_{max}$1 = 5, CI 2 = 0, CR 3 = 0

Bi (i = 1, …, 5) is the indicator of objective layer. B1—indoor micro-environment; B2—physiology of vegetable; B3—morphology of vegetable; B4—yield of vegetable; B5—quality of vegetable. ¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

(2)

Criteria layer B1

The criteria layer B1 includes four indicators: temperature C1, CO₂ concentration C2, humidity C3, illumination intensity C4. Ahmed [56] quantified the effects of CO₂ and illumination intensity on lettuce weight as 34% and 32%, respectively, indicating that CO₂ is slightly more important than light intensity. Zhou [57] investigated the interactive effects of light and temperature on photosynthesis and concluded that temperature exerts a more pronounced influence on both photosynthetic efficiency and lettuce productivity. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 4.1871, CR = 0.0693 < 0.1. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final B1-Ci weights are shown in

Table 6. Comparison matrix of indoor micro-environment (B1).

B1	C1	C2	C3	C4	${\mathit{\omega }}_1$
C1	5	1	3	5	0.58
C2	3	${\displaystyle \frac{1}{5}}$	1	1	0.17
C3	1	${\displaystyle \frac{1}{3}}$	1	1	0.15
C4	1	${\displaystyle \frac{1}{5}}$	1	${\displaystyle \frac{1}{3}}$	0.1
${\lambda }_{max}$1 = 4.1871, CI 2 = 0.0624, CR 3 = 0.0693

Ci (i = 1, …, 4) is the indicator of indoor micro-environment. C1—temperature; C2—CO₂ concentration; C3—humidity; C4—illumination intensity.¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

(3)

Criteria layer B2

The criteria layer B2 includes four indicators: root vitality C5, total chlorophyll C6, chlorophyll a C7, chlorophyll b C8. Roots are critical organs for water and nutrient uptake in plants, directly deciding both stress tolerance and growth potential [58]. Insufficient root vitality can limit plant growth, even when photosynthetic efficiency is high, thereby underscoring the vital role of root vitality in supporting healthy lettuce growth. Total chlorophyll content reflects the total amount of chlorophyll a and b and is an important indicator of photosynthetic capacity, which is more important than chlorophyll a or b alone. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 4.0455, CR = 0.0168 < 0.1. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final B2-Ci weights are shown in

Table 7. Comparison matrix of physiology of vegetable (B2).

B2	C5	C6	C7	C8	${\mathit{\omega }}_2$
C5	5	5	3	1	0.55
C6	3	3	1	${\displaystyle \frac{1}{3}}$	0.25
C7	1	1	${\displaystyle \frac{1}{3}}$	${\displaystyle \frac{1}{5}}$	0.1
C8	1	1	${\displaystyle \frac{1}{3}}$	${\displaystyle \frac{1}{5}}$	0.1
${\lambda }_{max}$1 = 4.0455, CI 2 = 0.0152, CR 3 = 0.0168

Ci (i = 5, …, 8) is the indicator of physiology of vegetable. C5—root vitality; C6—total chlorophyll; C7—chlorophyll a; C8—chlorophyll b. ¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

(4)

Criteria layer B3

The criteria layer B3 includes five indicators: stem diameter C9, number of blades C10, max leaf area C11, plant height C12, plant breadth C13. Photosynthesis is the fundamental physiological process through which plants convert inorganic compounds into organic matter, directly affecting the functional suitability of ILWs. Stem diameter serves as a critical indicator of a plant’s capacity for nutrient transport and resistance to lodging, thereby playing a pivotal role in maintaining the structural stability of the ILWs. Although the number of blades and maximum leaf area are also important contributors to photosynthetic efficiency, their influence is slightly less significant than that of stem diameter. Plant height and plant breadth, as primary morphological parameters, shape the overall appearance of vegetables but have a comparatively limited impact on photosynthetic efficiency. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 5.0778, CR = 0.0173 < 0.1. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final B3-Ci weights are shown in

Table 8. Comparison matrix of morphology of vegetable (B3).

B3	C9	C10	C11	C12	C13	${\mathit{\omega }}_3$
C9	3	4	1	2	3	0.4
C10	2	3	${\displaystyle \frac{1}{2}}$	1	1	0.21
C11	2	3	${\displaystyle \frac{1}{3}}$	1	1	0.2
C12	1	2	${\displaystyle \frac{1}{3}}$	${\displaystyle \frac{1}{2}}$	${\displaystyle \frac{1}{2}}$	0.12
C13	${\displaystyle \frac{1}{2}}$	1	${\displaystyle \frac{1}{4}}$	${\displaystyle \frac{1}{3}}$	${\displaystyle \frac{1}{3}}$	0.07
${\lambda }_{max}$1 = 5.0778, CI 2 = 0.0194, CR 3 = 0.0173

Ci (i = 9, …, 13) is the indicator of morphology of vegetable. C9—stem diameter; C10—number of blades; C11—max leaf area; C12—plant height; C13—plant breadth. ¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

(5)

Criteria layer B4

The criteria layer B4 includes two indicators: shoot fresh weight C14, root fresh weight C15. As a leafy vegetable, lettuce derives its economic value primarily from its edible foliage, while the root system is non-edible. Therefore, the fresh weight of the leaf parts is significantly higher than that of the root parts. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 2, CR = 0. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final B4-Ci weights are shown in

Table 9. Comparison matrix of yield of vegetable (B4).

B4	C14	C15	${\mathit{\omega }}_4$
C14	1	5	0.83
C15	${\displaystyle \frac{1}{5}}$	1	0.17
${\lambda }_{max}$1 = 2, CI 2 = 0, CR 3 = 0

Ci (i = 14, 15) is the indicator of yield of vegetable. C14—shoot fresh weight; C15—root fresh weight. ¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

(6)

Criteria layer B4

The criteria layer B5 includes three indicators: vitamin C content C16, soluble sugar content C17, nitrate nitrogen content C18. Vitamin C and soluble sugar content are positive indicators of nutritional quality, both of which directly influence the taste and nutritional value of lettuce and are equally valued by consumers. In contrast, although nitrate nitrogen is a negative safety indicator, its concentration in fresh lettuce is typically very low and has minimal impact on overall lettuce quality. Therefore, the importance of the first two is significantly higher than that of the latter. A pairwise comparison matrix was constructed accordingly, and the analysis results showed that ${\lambda }_{max}$ = 3.0057, CR = 0.0049 < 0.1. Having passed the consistency check, the resulting weighting distribution was deemed valid. The final B5-Ci weights are shown in

Table 10. Comparison matrix of quality of vegetable (B5).

B5	C16	C17	C18	${\mathit{\omega }}_5$
C16	1	1	5	0.47
C17	1	1	4	0.43
C18	1/5	1/4	1	0.1
${\lambda }_{max}$1 = 3.0057, CI 2 = 0.0028, CR 3 = 0.0049

Ci (i = 16, …, 18) is the indicator of quality of vegetable. C16—vitamin C content; C17—soluble sugar content. C18—nitrate nitrogen content. ¹ refers to the maximum eigenvalue. ² refers to consistence index (CI). ³ refers to consistence ratio (CR). CR < 0.1 passed the consistency test, the resulting weighting distribution was deemed valid.

.

3.2.2. Entropy Weights

Based on the experimental data for each indicator, the final entropy weights were calculated according to the above steps and Equations (6)–(9), as shown in

Table 11. Entropy weights.

Objective Layer	Criteria Layer	$\mathit{\omega }$	Indicator Layer		${\mathit{\omega }}_{\mathit{J}}$
Suitability evaluation of indoor living walls	Indoor micro-environment B1	0.20	Temperature C1	0.19	0.04
			CO2 concentration C2	0.20	0.04
			Humidity C3	0.23	0.05
			Illumination intensity C4	0.38	0.08
	Physiology of vegetable B2	0.20	Root vitality C5	0.25	0.05
			Total chlorophyll C6	0.23	0.05
			Chlorophyll a C7	0.33	0.07
			Chlorophyll b C8	0.19	0.04
	Morphology of vegetable B3	0.20	Stem diameter C9	0.14	0.03
			Number of blades C10	0.11	0.02
			Max leaf area C11	0.25	0.05
			Seeding height C12	0.24	0.05
			Plant breadth C13	0.26	0.05
	Yield of vegetable B4	0.20	Shoot fresh weight C14	0.57	0.11
			Root fresh weight C15	0.43	0.09
	Quality of vegetable B5	0.20	Vitamin C content C16	0.47	0.09
			Soluble sugar content C17	0.27	0.05
			Nitrate nitrogen content C18	0.26	0.05

.

3.2.3. Combination Weights

Based on the AHP weights (Table 9) and entropy weights (Table 10) calculated, the combination weights were calculated using Equation (10), and the results are shown in

Table 12. Combination weights.

Indicator	${\mathit{\omega }}_{\mathit{i}}$	${\mathit{\omega }}_{\mathit{J}}$	W	Indicator	${\mathit{\omega }}_{\mathit{i}}$	${\mathit{\omega }}_{\mathit{J}}$	W
C1	0.580	0.038	0.04	C10	0.042	0.022	0.01
C2	0.170	0.04	0.01	C11	0.040	0.05	0.03
C3	0.150	0.046	0.02	C12	0.024	0.048	0.02
C4	0.10	0.076	0.1	C13	0.014	0.052	0.01
C5	0.110	0.05	0.09	C14	0.166	0.114	0.29
C6	0.050	0.046	0.04	C15	0.034	0.086	0.05
C7	0.020	0.066	0.02	C16	0.094	0.094	0.14
C8	0.020	0.038	0.01	C17	0.086	0.054	0.07
C9	0.080	0.028	0.03	C18	0.020	0.052	0.02

.

3.3. Suitability Scores

Based on the methods described in Section 2.4 and Section 2.5, the measured data for all indicators in this study are summarized in

Table 13. Measurement data.

Indicators	R1-A	R1-B	R2-A	R2-B	R3-A	R3-B	R4-A	R4-B	F	p
Temperature C1 (°C)	19.5 ± 0.79	20.4 ± 0.91	22.17 ± 1.19	23.3 ± 1.12	24.75 ± 2.5	25.05 ± 2.62	25.5 ± 1.68	26.59 ± 1.8	6.613	<0.001
CO2 concentration C2 (ppm)	586.15 ± 1.64	579.36 ± 1.18	704.73 ± 3.34	706.3 ± 1.51	677.85 ± 1.98	679.97 ± 1.45	695.06 ± 1.7	699.73 ± 26.82	36.46	<0.001
Humidity C3 (%)	66.64 ± 2.71	61.05 ± 2.98	72.94 ± 7.9	71.44 ± 8.31	55.8 ± 13.31	54.76 ± 13.92	48.97 ± 3.1	46.24 ± 2.78	4.473	<0.05
Illumination intensity C4 (lx)	1100.02 ± 2.39	1119.22 ± 1.35	1141.41 ± 2.92	1155.08 ± 1.78	1363.35 ± 2.18	1393.68 ± 1.08	1550.42 ± 4.21	1585.36 ± 3.41	18.125	<0.001
Root vitality C5 (μg/h/g FW)	32.38 ± 0.97	39.58 ± 1.53	55.91 ± 2.05	63.34 ± 2.62	74.42 ± 1.47	80.22 ± 2	86.3 ± 0.84	91.14 ± 3.54	33.179	<0.001
Total chlorophyll C6 (mg/g FW)	0.61 ± 0.04	0.72 ± 0.04	0.84 ± 0.08	0.9 ± 0.08	1.05 ± 0.06	1.02 ± 0.04	1.09 ± 0.05	1.24 ± 0.08	34.609	<0.001
Chlorophyll a C7 (mg/g FW)	0.39 ± 0.03	0.4 ± 0.01	0.45 ± 0.05	0.47 ± 0.05	0.58 ± 0.06	0.58 ± 0.01	0.6 ± 0.02	0.68 ± 0.03	22.088	<0.001
Chlorophyll b C8 (mg/g FW)	0.22 ± 0.01	0.32 ± 0.03	0.39 ± 0.01	0.43 ± 0.03	0.48 ± 0.03	0.42 ± 0.05	0.49 ± 0.03	0.56 ± 0.06	28.354	<0.001
Stem diameter C9 (mm)	10.98 ± 0.09	11.85 ± 0.88	29.48 ± 0.66	31.6 ± 1.22	17.33 ± 0.47	18.15 ± 0.96	13.28 ± 0.48	14.63 ± 0.89	38.704	<0.001
Number of blades C10	6 ± 0.41	7 ± 0.58	9 ± 0.58	10 ± 0.58	8 ± 0.58	8 ± 0.58	7 ± 1.15	8 ± 0	11.032	<0.001
Max leaf area C11 (cm2)	33.78 ± 1.43	36.25 ± 1.52	158.97 ± 4.15	165.41 ± 3.78	93.73 ± 5.74	97.05 ± 2.78	50.14 ± 0.95	57.96 ± 4.48	67.773	<0.001
Seeding height C12 (cm)	9.9 ± 0.53	11.95 ± 0.84	27.58 ± 0.52	29.53 ± 1.08	16.66 ± 1.05	17.38 ± 0.49	11.63 ± 0.8	13.78 ± 1.05	28.262	<0.001
Plant breadth C13 (cm)	10.98 ± 1.08	11.85 ± 1.6	29.48 ± 1.51	31.64 ± 1.41	17.33 ± 0.95	18.15 ± 1.72	13.28 ± 0.91	14.63 ± 0.77	13.161	<0.001
Shoot fresh weight C14 (g)	33.78 ± 1.27	36.25 ± 1.09	158.98 ± 1.41	165.41 ± 1.63	93.73 ± 1.24	97.05 ± 1.72	50.14 ± 1.55	57.96 ± 1.64	47.947	<0.001
Root fresh weight C15 (g)	8.72 ± 0.39	9.48 ± 0.4	35.83 ± 0.87	39.07 ± 2.39	16.42 ± 0.8	19.47 ± 1.42	10.13 ± 0.73	12.1 ± 1.7	20.906	<0.001
Vitamin C content C16 (mg/g)	0.08 ± 0.02	0.08 ± 0.01	0.11 ± 0.02	0.12 ± 0.04	0.07 ± 0.02	0.07 ± 0.01	0.07 ± 0.02	0.08 ± 0.02	27.155	<0.001
Total soluble sugar content C17 (mg/g)	5.08 ± 0.16	5.85 ± 0.12	8.48 ± 0.15	8.75 ± 0.05	7.19 ± 0.11	7.59 ± 0.02	6.58 ± 0.09	6.13 ± 0.08	42.08	<0.001
Nirate nitrogen C18 (mg/kg)	223.77 ± 1.85	212.25 ± 2.54	148.69 ± 1.43	143.75 ± 0.92	133.72 ± 0.61	145.03 ± 1.58	110.49 ± 2.05	104.86 ± 0.66	69.095	<0.001

Statistical analysis of the experimental data was performed using SPSS 27.0 software. A one-way ANOVA analysis was used to analyze significant differences, and the results showed significant differences (p < 0.05) between groups (n = 3). The data in the table are presented as “mean ± standard deviation.”

. By comparing the measured data in Table 13 with the range of criteria for each indicator grade in Table 4, values were assigned to the 18 indicators C1 to C18. The final assignment results are detailed in

Table 14. Assignment value.

Group	C1	C2	C3	C4	C5	C6	C7	C8	C9	C10	C11	C12	C13	C14	C15	C16	C17	C18
R1-A	2	2	2	1	1	1	1	1	1	1	1	1	1	1	1	1	1	3
R1-B	2	2	2	1	1	1	1	1	1	2	1	1	1	1	1	1	1	3
R2-A	3	2	3	1	1	1	1	1	2	3	3	3	2	3	2	1	2	3
R2-B	3	2	3	1	2	2	1	2	3	3	3	3	3	3	3	1	2	3
R3-A	3	2	2	1	2	2	2	2	2	2	2	2	2	1	2	1	1	3
R3-B	1	2	2	1	2	2	2	2	2	2	2	2	2	2	2	1	1	3
R4-A	1	2	2	2	3	3	2	3	1	2	1	1	1	1	1	1	1	3
R4-B	1	2	2	2	3	3	3	3	1	2	1	1	1	1	1	1	1	3

Compare the measured data with the data range in Table 4 to determine its score.

.

Based on the combination weights presented in Table 12, the evaluation scores for each experimental group were calculated using Equation (11), with the specific results summarized in

Table 15. Suitability scores.

Group	R1-A	R1-B	R2-A	R2-B	R3-A	R3-B	R4-A	R4-B
Score	1.15	1.17	2.17	2.41	1.55	1.72	1.42	1.44

The study totally sets eight experimental sample groups (R1A~R4-B). Calculated by using Equation (11) and values in Table 14 to scores.

. The data suggest that the suitability of ILWs is significantly influenced by spatial layout, with the R2 configuration demonstrating the most favorable performance. Groups A and B received scores of 2.17 and 2.41, respectively, both of which fall under the “suitable” category. The remaining groups, ranked from highest to lowest, are R3-B, R3-A, R4-B, R4-A, R1-B, and R1-A; all are classified as “moderately suitable.” In addition, planting height appears to affect ILWs suitability, with a consistent trend observed: in every experimental unit, Group B outperformed Group A. Under the climatic conditions of Hainan Province, the optimal spatial configuration for ILWs involves north-facing lighting paired with a west-wall planting layout. The recommended installation height ranges from 1.3 to 1.75 m (corresponding to Group B height), while installation at lower heights should be avoided.

4. Discussion

4.1. Environmental Suitability

The environmental indicator measurement data and corresponding weights used in this study are presented in Table 13 and Table 14. Analysis reveals that among various indoor microenvironmental factors, temperature holds the highest weighting in terms of its influence on lettuce growth. The recorded data indicate that the temperature C1 increased progressively from R1-A to R4-B, with rooms facing south (R3 and R4) exhibiting generally higher temperatures compared to those facing north (R1 and R2). This variation is primarily attributed to the extended exposure to sunlight in south-facing rooms, where direct solar radiation leads to continuous heat accumulation. Notably, even during winter, the temperature in R4 exceeded 25 °C, surpassing the optimal thermal range for lettuce cultivation. Previous studies have demonstrated that prolonged exposure to high temperatures can result in slow growth and excessive stem elongation in lettuce, significantly compromising its quality [59]. Therefore, in region of Hainan Province, maintaining consistent control of indoor temperature is crucial for the successful operation of vertical farming systems. It is recommended that the cooling capacity of such environments be appropriately increased to meet this requirement.

The measured values for illumination intensity C4 also demonstrated a gradual increase from R1-A to R4-B, with only the R4 group meeting the minimum light compensation point required for optimal lettuce growth. Significant differences in natural light exposure were observed between planting walls at different heights, with Group A generally exhibiting lower light intensity than Group B. This disparity is primarily attributed to multiple obstructions—such as furniture, window frames, and upper planting slots—that impeded light penetration to the lower levels. Liu [60] confirmed that in natural light-based plant factories, the lower layers in multi-tiered planting systems experience substantial shading. Consequently, the integration of artificial lighting is essential to improve the light environment, ensure uniform light distribution, and ultimately enhance the quality of lettuce cultivation.

Humidity test data indicate that relative humidity levels in north-facing rooms are significantly higher than those in south-facing rooms. This difference is primarily attributed to prolonged sunlight exposure in south-facing rooms, which increases temperature and accelerates water evaporation, thereby reducing relative humidity. Future designs of ILWs may consider coordinated integration with a building’s fresh air ventilation system. Such integration would not only meet the CO₂ requirements for plant photosynthesis but also utilize CO₂ emissions from buildings as a carbon resource, thus enabling synergistic optimization across energy efficiency, environmental regulation, and agricultural productivity. In summary, when installing indoor living walls in region of Hainan Province, the following considerations should be prioritized: spatial layouts should favor west-facing walls within north-oriented rooms; vertical installation height should exceed 1.3 m; and supplemental artificial lighting and temperature regulation systems must be incorporated.

4.2. Vegetable Suitability

This study assessed multiple dimensions of lettuce development and final product quality to evaluate vegetable suitability. Indicators C5 through C8 represent the dynamic growth processes of lettuce, while indicators C9 through C18 reflect its final morphological and physiological characteristics. Vegetable physiological indices are indicative of photosynthetic efficiency and metabolic activity during the vegetative growth phase, serving as the fundamental drivers of yield formation. As shown in Table 12, the weight calculation results identify root vitality C5 as the most influential factor among all vegetable physiological variables, indicating its dominant role in lettuce growth performance. According to Table 14, root vitality and chlorophyll content were lower in the north-facing rooms (R1 and R2) compared to the south-facing rooms (R3 and R4). Root vitality is strongly influenced by both substrate conditions and temperature [61]. Given that all experimental groups employed a standardized substrate composition, variations in root vitality cannot be attributed to substrate differences. Consequently, temperature emerged as the primary influencing factor, exhibiting a positive correlation with root system activity—higher temperatures were associated with enhanced root vitality. This suggests that higher temperatures facilitate organic matter synthesis; however, excessive heat may have adverse effects, thereby emphasizing the necessity of environmental control systems for continuous monitoring and regulation in applications of ILWs.

In terms of morphological performance, the R2 experimental unit outperformed all other groups. As shown in Table 13, the R2-B group achieved 3 points for all key morphological indicators, including stem diameter C9, number of blades C10, maximum leaf area C11, plant height C12, and plant breadth C13, demonstrating superior lettuce morphology. Lettuce morphology is predominantly influenced by light conditions. Iqbal [62] confirmed that under stable photoperiod conditions, increased illumination intensity significantly enhances morphological traits such as plant height, plant breadth, and leaf count. Additionally, extending the photoperiod under a fixed light intensity also contributes to improvements in these characteristics. It is important to note that this study was conducted in a typical office environment. Although R3 and R4, which face south, receive strong natural light, the need to maintain visual comfort for office occupants often necessitates the use of curtains during working hours, thereby reducing light exposure and leading to insufficient photoperiods for the plants. This instability in light availability likely contributed to the suboptimal morphological outcomes observed in R3 and R4. Furthermore, as conventional office spaces, R3 and R4 experience frequent human activity, which may have interfered with plant growth. In contrast, R2—used primarily as a meeting room—has a lower occupancy rate and offers more stable environmental conditions, potentially contributing to its superior morphological results. These findings suggest that ILWs should adopt a “human–plant semi-isolation” spatial layout, wherein partial separation within shared spaces enables a balanced coexistence between plant development and human activities.

In terms of vegetable yield, the yield in this study was slightly lower than that of professional plant factories. Among all experimental groups, R2 demonstrated the highest performance, with fresh weights of 39.1 g and 35.8 g in Groups B and A, respectively. The fresh weights for the remaining groups were as follows: R3-B (19.470 g), R3-A (16.423 g), R4-B (12.100 g), R4-A (10.133 g), R1-B (9.480 g), and R1-A (8.715 g). According to data from the Chinese National Bureau of Statistics and the World Health Organization (WHO), the recommended daily intake of fresh vegetables per adult ranges between 400 and 800 g. However, the actual average per capita vegetable consumption in China is 109.8 kg per year, equivalent to approximately 300 g per day. For a typical five-person office, the required daily supply would be between 2000 and 4000 g. In contrast, the maximum daily yield recorded by the experimental ILWs unit was only 41.195 g, which is insufficient to meet full dietary demand. However, the ILWs adopted in this research is relatively concise, and the output has been somewhat limited. The existing unmanned plant factory can produce 180 kg of vegetables per day [63]. Future improvements in planting density, system design, and automated environmental control could significantly enhance production capacity. These findings support the feasibility of ILWs as a supplementary vegetable supply for office environments and provide a clear direction for increasing yields through technological innovation.

This study evaluated vegetable quality based on three key parameters: vitamin C content C16, soluble sugar content C17, and nitrate nitrogen content C18, which directly reflect the commercial value and edibility of lettuce. The biosynthesis of vitamin C is strongly influenced by ambient temperature, and prior research has shown that elevated temperatures inhibit ascorbic acid accumulation [64]. The findings of this study align with those results: as indoor temperatures increased from Group R2 to Group R4, the vitamin C content exhibited a notable downward trend. Soluble sugars, which enhance both the flavor and nutritional value of lettuce, are closely tied to light [65]. Although south-facing rooms R3 and R4 received higher natural light intensity, their soluble sugar content was slightly lower than that of the R2 group, likely due to light instability caused by frequent human activities. Previous studies have indicated that short-term continuous lighting prior to harvest can enhance soluble sugar accumulation in lettuce [66]. This suggests that regulating light conditions during the pre-harvest period is particularly important for improving lettuce quality in ILWs. Importantly, nitrate levels in all experimental groups were well below internationally accepted food safety thresholds, confirming that lettuce produced by ILWs is safe for consumption. In summary, further optimization of temperature regulation and light management has the potential to significantly improve the nutritional quality and flavor profile of crops grown in indoor living walls environments.

4.3. Implications for Sustainability

The suitability evaluation model serves as a critical method linking the promotion and application of BIA with urban sustainability goals. By scientifically and systematically quantifying the suitability of ILWs, this model provides a theoretical foundation for their practical implementation, thereby achieving multifaceted environmental, social, and economic benefits.

From an environmental perspective, appropriately designed ILWs can significantly improve indoor environmental quality and reduce building energy consumption through the physiological activities of plants. Previous studies have shown that, regardless of whether ornamental or edible plants are used, ILWs can play an important role in enhancing indoor air quality and improving energy efficiency in buildings. Ornamental species such as Sansevieria trifasciata and Epipremnum aureum can reduce fresh air demand by 13.9~38.5% and lower fresh air energy consumption by 11.2~28.2% [16]. Similarly, introducing 100 lettuce plants into an office of approximately 30 m² was found to decrease indoor CO₂ concentration by 25.7~34.3% and reduce building ventilation energy consumption by 12.7~58.4% [20].

From a social perspective, the ILWs provide more opportunities for people to connect intimately with nature, contributing to improved mental health and reduced anxiety levels [19]. For instance, at Pasona company’s indoor farm operations, gardening activities fostered communication among employees and strengthened collective identity.

In economic terms, suitability assessment models help enhance the production efficiency, operational reliability, and service life of ILWs, thereby ensuring their long-term stable operation and sustainable development. Compared to high-tech, high-cost plant factories, simple, low-cost ILWs are more readily scalable and applicable. Simultaneously, localized production not only reduces logistics and preservation costs but also provides urban residents with green, healthy, and sustainable food sources, thereby driving innovation and development in urban agricultural economic models. This approach also helps shorten food miles, reducing carbon emissions and energy consumption associated with transportation and storage.

5. Conclusions

This study combined the AHP with the entropy weight method to construct a suitability assessment model for indoor living walls (ILWs), effectively overcoming the limitations of computationally complex fuzzy mathematical methods and the highly subjective Delphi method. The model can scientifically evaluate the suitability of ILWs for growth in indoor environments with different spatial characteristics, providing a quantitative basis for system site selection, plant configuration, and operation and maintenance strategies.

The study revealed through a comparative assessment of four typical spatial layouts that the “human–vegetable semi-isolation” spatial model combined with north-facing lighting and west-wall planting layout has the most optimal suitability. This finding not only reveals the key influence of spatial configuration on the effectiveness of ILWs but also provides theoretical support and practical guidance for the advancement of BIA design. However, this study has limitations because it does not include a systematic analysis of energy efficiency and water resource management. In practical applications, the use of water and energy should be considered in detail. In addition, limited automation of experimental facilities has resulted in restricted production capacity. This can be addressed by increasing the automation of ILWs and improving environmental control.

The future development of ILWs should prioritize the advancement of intelligent environmental control systems that integrate real-time monitoring and automatic regulation of key parameters, including temperature, illumination intensity, CO₂ concentration and so on. These technologies are essential for establishing high-efficiency production systems. Emphasis should also be placed on optimizing the “human–plant semi-isolation” spatial layout model and planting height range. Furthermore, the development of high-density vertical planting modules and management modes is necessary to improve both crop yield and quality, thereby advancing the goal of self-sufficiency in BIA. At the same time, the introduction of building information modeling (BIM) technology can be considered for planting scheme simulation and optimization, enabling the low-cost, high-efficiency, large-scale application of ILWs.