Life cycle analysis guides decision making and policy for CEA (Fig. 1). Early-stage assessment across the life cycle of CEA is essential to support integrated decision-making and minimize costs. Life cycle analysis can be used to optimize key CEA design factors, such as CEA size, location, envelope design, and heating, ventilation and air conditioning (HVAC) systems, and guide research and development to identify critical technologies for the CEA industry. However, there is a lack of comprehensive life cycle analysis, evaluating potential environmental, economic and social impacts of the new CEA designs at different settings.
Most existing CEA life cycle analyses, such as life cycle assessment (LCA), evaluate carbon footprints of resource use (e.g. fertilizer, energy, water) or both economic and environmental performance through case studies. Linking LCA to a bioeconomic model has been proposed for analyzing area-based farming policy. Most studies represent current CEA technologies and have well-defined CEA in terms of size, system, and location. Thus, these studies are limited to comparing the current CEA prototypes to open-field agriculture but have limited impacts on supporting new designs and decision making for CEA.
Advocates for an ecological-economic approach that considers the environmental dimensions of meeting human needs highlight the need for a comprehensive approach to LCA that integrates economic, social, and environmental aspects. Weidema emphasizes the importance of integrating social aspects into LCA, while Arodudu et al. suggests the use of tools and methodologies to bridge methodological gaps in LCA application to agro-bioenergy systems and to integrate agronomic options and life cycle thinking approaches.
LCA can be further applied to fulfill the CEA circular economy, supporting the development of sustainable strategies for CEA industry, reduction in waste and cost, and enhancement of resource efficiency at community scales. The reuse and recycling opportunities include waste heat utilization, CO supply through co-location, water reuse and reclamation of nutrients from water treatment plants, and recycling of growing media and food packages. Various case studies report their utilization of waste heat from combined heat and power plants, data centers, and plant factories. Indoor farms reuse low-quality energy, in the form of warm water or air in a temperature range of 30-47 °C. Recirculating irrigation water has the potential to reduce water consumption by 20-40% and fertilizer costs by 40-50%. Water reuse for crop growth conserves freshwater, promotes water circularity, and reduces the total need for chemical fertilizers. Defining the water treatment and nutrient reclamation requirements for crop growth is an important research question for CEA. Potential contamination risks such as microbial pathogens and chemical content should also be considered for irrigation. Although waste resources may be limited by location, it is possible to create a CEA ecosystem through strategic planning of businesses and infrastructure within communities. The evaluation of the planning scenarios can be based on comprehensive life cycle analysis, integrating economic, social, and environmental aspects.
Distributed indoor agriculture (DIA) systems within buildings can support crop cultivation using hydroponic techniques -- such as NFT and DWC, as discussed in Section "Current CEA technologies". Each DIA system (Fig. 2) is equipped with integrated hardware and software solutions, including controls of grow lights and irrigation, monitoring of crop growth, and diagnosis of issues related to crop health and system operation.
Indoor living walls, a foundational prototype of the DIA system, have gained popularity by bringing nature indoors, and indoor plants have become highly desirable with the rise of remote work. Studies show measurable benefits of these systems on occupants' thermal comfort and building cooling load reduction. With living walls, the cooling setpoint can be increased by 0.9 K and still satisfy thermal comfort needs for the majority of the occupants. In an experimental study of a hall with a floor area of 520 ft that was not equipped with air conditioning, an average temperature reduction of 4 K was observed. In addition, indoor plants demonstrate the capability to improve indoor air quality, productivity and creativity, while reducing noise levels, the negative effects of visual glare, stresses, and discomfort symptoms. Plant leaves and their microbes purify indoor air through phytoremediation, absorbing pollutants via leaf stomata and degrading them through plant metabolism. Plant elements like leaves and twigs reflect, scatter, and attenuate sound through mechanical vibration. A study of a 2.4 m living wall in a space with floor area of 19.6 m found a weighted sound reduction index of 15 dB and a weighted sound absorption coefficient of 0.40. Moreover, people were more productive (12% quicker reaction time on a computer task) and less stressed (systolic blood pressure readings lowered by one to four units).
Can DIA systems be a potential integrated solution to food resilience, better indoor environmental quality, energy efficiency, and well-being? Broadly speaking, there are four broad markets to target (1) offices; (2) institutions; (3) hospitality; and (4) residential buildings. Dedicating 1% of floor area identified in the Commercial Building Energy Consumption Survey (CBECS) and Residential Energy Consumption Survey (RECS) to growing lettuce in existing buildings has a potential yield of 5.1-70.4 kg fresh weight lettuce per capita per growing cycle (about four weeks) based on the reported lettuce yield (6.9-95 kg fresh weight) per growing floor area (m). This potential yield is much higher than the U.S. per capita consumption of fresh romaine and leaf lettuce (5.8 kg) in 2022. DIA and CEA are distinct and potentially competitive models for food production, but their co-existence can provide mutual benefits. Successful DIA demonstrations promote mass-market acceptance of hydroponically grown foods through social impacts, public education, and easy accessibility of technology.
Connecting CEA with microgrids can not only reduce CEA's carbon footprint but also contribute to stabilizing microgrids through demand response and frequency control. Electricity demand from lights and HVAC systems and the associated carbon footprints have been significant for indoor vertical farms. CEA can be strategically co-located in areas with existing microgrid infrastructure or where dynamic electricity pricing creates opportunities for cost savings and load flexibility. Shifting LED lights from continuous to intermittent operation was implemented to reduce operational costs by responding to daily electricity price fluctuations.
CEA can serve as a demand-side dispatchable load for microgrids to reduce the volatility of renewable energy resources (Fig. 3). Stable operation of electric systems requires frequency regulation; providing frequency regulation from demand side resources mitigates technical, economic, and political challenges. Existing studies have tested the feasibility of using HVAC equipment as dispatchable loads on the demand side. Due to the high volatility of such distributed energy resources (DERs) as wind turbines, photovoltaics, and hydroelectricity, balancing renewable energy generation and demands for microgrids is expensive in comparison to traditional power grid frequency regulation. CEA can potentially serve as an important dispatchable load on the demand side if crops can tolerate certain lighting fluctuations. Dynamic environmental variations trigger plants' physiological response, yet they can maintain diurnal leaf carbon gain in comparison with plant growth in constant environments. Further research is needed on how variations of LED light and temperature induced by dynamic variations in price signals from microgrids affect crop yield and quality.
A CEA Digital Twin (DT) framework integrates computer vision, edge computing, AI-enabled predictive analytics, and optimal control technologies (Fig. 4), forecasting and optimizing the behavior of the physical asset (crops) and resource management. DT frameworks have been explored in various contexts in CEA, including monitoring DT for tracking subsystems, predictive DT for optimizing production and resource use efficiency, supply chain DT to optimize agrifood supply chain, DT for aquaponics production facilities, and visualization DT.
Computer vision is the key component to acquire crop growth information and provide feedback to update crop model parameters for resource optimization. Studies have focused on developing growth forecasting based on early-stage growth images using spatial transformation and spatial-temporal attention mechanisms. State-of-the-art algorithms, such as YOLO series, Mask-RCNN, and Deepabv3+, have been employed for growth monitoring, yield prediction, and spacing optimization. However, image-based systems encounter limitations. Occlusions and complex canopies create challenges in developing automation solutions for plant monitoring. Thus, advanced algorithms and approaches such as 3D reconstruction (via Neural Radiance Fields, and Gaussian Splatting), and geometric DT are needed to enhance crop growth monitoring solutions for CEA. Furthermore, multispectral and hyperspectral imaging systems have been deployed to determine the nutrient concentration in CEA produce so that growers can develop mitigation strategies before deficiency symptoms arise. However, the detection accuracy is limited due to the extremely complex workspace. Most deep learning-based networks require image normalization before images are input to the network, reducing the feature details and decreasing detection accuracy.
Intelligent climate controllers, driven by microclimate data, have advanced optimal energy use management. AI algorithms, such as reinforcement learning have been explored to optimize energy and water use in greenhouses. A wide range of control algorithms has been explored to optimize variables such as light intensity, air velocity, shade curtain, vapor pressure deficits, and CO levels to minimize energy costs. In addition, DT requests vast amounts of data, subject to issues like inaccuracies and cyberattacks. Blockchain has been applied in CEA applications to enhance information security.
A major obstacle to optimal outcomes in CEA lies in real-time crops' response to environmental conditions and thus decoupled crop growth and resource optimization. Although studies have been conducted for optimal resource allocation under fixed microclimate conditions, forecasting growth based on training data from fixed conditions may limit its applicability in real CEA environments with dynamic climate conditions. A real-time feedback DT system must assess forecast accuracy and adjust model parameters, leading to efficient decision support for growers to optimize production and energy use.
Growing plants in hydroponics rather than soil dramatically changes the challenges. One important facet of plant physiology and plant health is the contribution of the plant microbiota, which represents a largely untapped opportunity. Plant growth-promoting bacteria (PGPB) reside in or around plants and can act as biostimulants, biofertilizers, and bioprotectants. Plant-associated beneficial microorganisms - e.g., bacteria and fungi - promote growth, nutrient uptake, stress tolerance, and resistance to pathogens. However, many plant-growth-promoting microbes found in soil cannot make the transition to hydroponic environments. Moreover, many PGPB products are developed for field application and have not been optimized for specific conditions of CEA which include dense cropping of specifically optimized crops, controlled water and climate systems, and integrated sensors and controllers. While hydroponics has a range of advantages over soil-based cultivation, one of the disadvantages is the fast spread of infectious diseases due to the recirculating nature of nutrient solution in the whole system. In such a case, PGPB may offer a unique solution in hydroponic systems by sensing and preventing pathogen outbreaks, improving access to nutrients and tolerance to biotic and abiotic stresses and increasing crop yield. To date, there are reports that PGPB promotes plant growth in various hydroponic systems. For example, Pseudomonas psychrotolerans IALR632 increased shoot and root growth of green Oakleaf lettuce grown in nutrient film technique (NFT) in the greenhouse, indoor vertical NFT, and a deep-water culture system.
Another group of beneficial microbial-based microorganisms is arbuscular mycorrhizal fungi (AMF). AMF is known to form AMF-host plant symbiosis with more than 80% of plant species. The primary positive impacts of AMF symbiosis are increasing availability of both macro- and micro-nutrients, increasing photosynthetic rate, and enhancing tolerance to stressful conditions through augmentation of antioxidant defense system. A recent study by Caser et al. for saffron in soilless systems in a glasshouse indicated that inoculation with one AMF species (Rhizophagus intraradices) or a mixture of R. intraradices and Funneliformis mosseae increased spice quality as evidenced by a superior content of several health-promoting compounds (polyphenols, anthocyanins, vitamin C, and antioxidant activity) in one cycle of growth in soilless systems compared to open field production, while spice yield was similar to that of open field. These results improve our understanding of microbial communities in soilless media. Nevertheless, such information is still limited, and future studies are needed to fully understand and optimize the benefits of microbes under controlled environments. There are also advanced opportunities for genetically engineering these communities for enhanced properties for the plant; better association with the plant and survival/persistence in the hydroponic environment, and for coupling to CEA control systems by reporting on water, microbiome and plant health through expression of environmentally-sensitive fluorescent and volatile organic reporters. A growing body of literature has focused on defining, engineering and using laboratory evolution to evolve such synthetic communities for plants in both soil and artificial conditions -- a focused effort to co-design microbial communities supporting optimized plants and coupling with control systems.
Transgenic crops have revolutionized the scale and nature of agriculture; however, such efforts are largely focused on addressing producer-facing challenges associated with traditional methods of growing crops in fields (i.e., herbicide tolerance, insect resistance). Crops grown in CEAs may shift the focus of engineering efforts towards other consumer-facing traits that may benefit from indoor growth. Growing plants that are engineered to improve human health through the enhanced delivery of key phytonutrients has been a key pillar in plant engineering efforts. Complex natural products ranging from edible health compounds to plant-derived pharmaceuticals may be developed as interesting targets for transgenic plants grown in CEA. Nutrient management in CEA can significantly enhance the vitamin content in crops through precise control and optimization of nutrient solutions that meet the specific requirements of various plant species and growth stages. This targeted approach enables growers to enhance the uptake of specific nutrients that are precursors to vitamins, leading to increased vitamin content in crops. One regulatory challenge of field-grown crops is the concern of outcrossing of transgenes; thus, the large-scale indoor growth of transgenic crops in CEA may open the door to safer implementation and growth of engineered plants, which may also expand the range of traits that could be pursued.
Optimizing and redesigning plant architecture within confined space limitations has the opportunity to dramatically enhance yield of indoor-grown crops. Kwon et al. leveraged CRISPR-Cas9 genome editing to target and manipulate the physiology of tomato plants by engineering highly compact and rapid flowering plants. The optimization of plant stature may enable the custom development of CEAs that are more space efficient and enable higher yield per footprint. The implications of such efforts are aligned with the challenges associated with growing plants indoors with limited space.