A Rapid Fabrication Methodology for Payload Modules, Piloted for the Observation of Queen Honey Bees (Apis mellifera) in Microgravity

The private space sector is driving the unprecedented availability of modern microgravity and spaceflight experimental opportunities (Crawford, 2016). However, the feasibility of space-bound research remains limited by two factors: the need for custom modules and costs. Researchers have a limited capacity to fabricate modules that meet payload requirements as well as satisfy their specific experimental needs (Pletser and De Crombrugghe, 2017). Biological researchers face even higher barriers, as the design complexity and specialization necessary to manufacture custom modules for living organisms is an even greater challenge (Boggs and Dhiri, 2020). While services for subdivided payload platforms may offer more affordable slots, they require even further weight and volume limitations. Thus, improvements in module weight minimization and design simplicity would lead to a reduction in costs, expertise, and time necessary for payload development and would be incredibly valuable for microgravity research, including biological studies. The aim of this paper is to respond to this opportunity by providing a generalizable process for designing and fabricating custom, lightweight housing for experimental modules that foster minimal material sourcing, parts, and assembly.

The field of biological module design has made significant progress in improving accessibility. In contrast to previous initiatives, such as the ISS's 2001 Insect Habitat that demanded $2.7M and 2 years for preliminary development (Rock, 2002); more recent designs, such as the BioExplorer-inspired Fruit Fly Module (Inan et al., 2009) and laboratories customized for low-cost CubeSat platforms (Poghosyan and Golkar, 2017), have demonstrated increasingly efficient and versatile precedents. However, modules continue to rely on “frame and fixture” approaches, where components are individually affixed to an aluminum chassis, often through multiple custom-tailored fixtures and braces (Richter et al., 2010). This process requires intensive materials and parts sourcing, long turnaround times, and intricate assemblies. These challenges are further compounded by launch costs (e.g., ~$10,000/lb orbital payload fee) and payload specification and development factors (e.g., NanoRack's “2U” format constrains an experiment to 0.50 kg and 10 × 10 × 20 cm with a development time of (~6–12 months from slot confirmation). We present a new methodology to address the challenges of both the complexities of custom experiment module design and the general payload specifications (cost of launch, weight, and dimensions) for wide-ranging experimental subject matter and function.

Our model case used in this work is an important and understudied area of space biology: the effect of microgravity on the health and reproductive viability of honey bees (Apis mellifera). Honey bees are an integral part of food production, responsible for pollinating 80% of edible flowering crops and enhancing the nutrition, yield, and shelf-life of produce beyond hand-pollination (Stein et al., 2017). These properties make honey bees promising components for supporting the regenerative food systems considered necessary for long-term space missions (Anderson et al., 2015). Importantly, one queen bee represents the laying potential of 1.3 million bees, with organs to store both male and female germlines (Nolan, 1925). A queen's fitness is a direct predictor of colony health and productivity, while reciprocally, a retinue of approximately 10 worker bee attendants is crucial for maintaining the queen's base metabolic regulation (e.g., food intake, temperature homeostasis, and excretion removal).

Tracking queen movement via behavioral observations has been the focus of previous research seeking to understand the impacts of stressors on queen and subsequent colony health (Wu-Smart and Spivak, 2016). The retinue response is also an established metric of queen quality, because the number of attendants that are attracted to a queen is largely determined by her pheromone production profile, which in turn is influenced by mating quality, reproductive status, and external stressors; thus, retinue response represents an honest indicator of a queen's fitness (Margarita et al., 2020; Rangel et al., 2016; Richard et al., 2007). Also, recent studies on the effect of temperature during long-distance transportation in migratory beekeeping have recommended that vibration, noise, and atmospheric pressure are also potential sources of stress that may contribute to colony loss (Jara et al., 2021; Melicher et al., 2019). Yet, beyond the observations of the honey bees aboard STS-13 (Vandenberg et al., 1985), which exhibited inviable eggs post-mission, no research exists on queen bee quality and fecundity in spaceflight environments.

Our work examines the intersection of these problem areas. To ease the current bottleneck in biological space research, we developed a low-cost, rapid design and fabrication methodology for modules compatible with commercial payload slots. In the Methods section, we present a computer-aided design (CAD) and subtractive manufacturing process for fabricating a custom module housing that utilizes a single low-density, rigid foam and occupies the payload's entire volume sans cavities for functional components. In addition, we detail a lightweight, video-enabled module using this design methodology to perform an original observational study on queen bees onboard the Blue Origin New Shepard 11 (NS-11) suborbital flight. The Results section documents our module's performance in structural evaluations, weight reduction, and ease of assembly of experimental components. Further, we analyze the video footage from the NS-11 flight to determine the regulatory capacity of honey bees in various stages of spaceflight. The data are presented as illustrative examples of data feasible on this platform, yet also may be informative to our understanding of spaceflight stressors relevant to queen bee health. Finally, we discuss the significance of this methodology's ability to address the current limitations in module design and to capture data on previously limited areas of space biology.

First, the results provide an evaluation of the module design through: (i) validation of performance to technical standards, (ii) demonstration of weight reduction compared to traditional methods, and (iii) ease of assembly and installation of living specimens (Figure 1). Second, the results demonstrate biological experimental success via autonomous support and monitoring of honey bee specimens during the NS-11 flight and the generation of new data on queen-worker behavior in microgravity (Figure 2).

Figure 2

For validation of performance to technical standards, the module design methodology (Figure 1A) successfully demonstrated a new type of lightweight foam chassis that acted as a multidirectional housing for all functional components while also providing shock absorption, thermal and electrical insulation, and fire retardation. Payload specifications require housing to withstand nominal and non-nominal static and dynamic forces, serve as a physical enclosure, and protect all experimental components from vibrations, shifting, or becoming unanchored during flight. Our chassis comprised a rigid polyurethane foam (FR-7106, General Plastics) machined in a split-form that snugly held each internal component in place and additionally offered vibrational dampening and electrical insulation. In Table 1, the chassis satisfied a panel of payload tests requirements for static and dynamic load (≤30 lb force), vibration (7.85 g at 50 Hz frequency), heat transfer (non-significant), flame (self-extinguishing), and dust formation (non-significant).

In demonstration of weight reduction, our 2U-form-factor chassis (machined from a foam density of 96 kg/m³ and coated in a protective paint) weighed 0.13 kg. Importantly, our chassis also functioned holistically as the frame, fixtures, and walls, leaving all but 26% of payload weight to the experiment. By contrast, existing 2U payloads have achieved a bare-frame weight of 0.16–0.28 kg (32–55% of the total payload weight) not including the required fixture hardware or wall panels (NASA, 2018b).

Regarding ease of assembly, our method enabled toolless and reversible press-fit assembly of all experimental components, which proved markedly beneficial to the rapid loading and retrieval of biological samples. Illustratively, we generated a platform for conducting an autonomous biological experiment for inclusion in a NanoRacks Feather Frame subdivided payload for suborbital spaceflight. The completed assembly, detailed in Figure 1B and Table 2, comprised over 24 interrelated components organized to initiate and perform stabilized video monitoring of an experiment chamber, closed-loop temperature control, and data logging from internal sensors and flight telemetry. At handoff (L-30 days), the housing was closed via press-fit, guided by four alignment pins, and fastened with two silicone bands in machined grooves (Figure 1D–F); during a late load, hours prior to launch (L-0 days 7 h), the assembly was easily reopened without tools and honey bee samples were rapidly loaded.

Finally, in demonstration of biological experiment execution, two space-bound modules performed an autonomous experimental program during spaceflight: each monitored internal temperature and humidity continuously and recorded 18.75 min of countdown-triggered audiovisual data of subjects S1 and S2 (2.7K/2704 × 1520 pixels, 60 fps, stabilized; Figure 2A, SI Video 1). The module experienced a 3.1-minute micro-gravitational arc (<0.1 G) and up to 6.5 G during descent. Results used visual data to analyze queen-worker capacity for homeostatic regulation under microgravity and flight stress, compared to grounded controls C1 and C2. Consistent with prevailing entomological methods to determine queen bee health (Slessor et al., 1988; Vergoz et al., 2009), video data was first analyzed for bee spatial distribution (Figure 2B,C) and queen tracking (Figure 2D). Then, as a proxy for retinue formation, we derived retinue area (cluster area per timepoint) and queen distance (distance of the queen from on the mean center of the cluster). Figure 2E and F pairs flight data (velocity, acceleration, altitude, flight stage) with the retinue area and queen distance of subjects S1, S2, and C1.

We used the queen distance measurement (Figure 2F, orange line), in conjunction with an established definition for the active retinue ellipse as a ≤1.5 cm distance from the queen (Kaminski et al., 1990, orange dashed line), to establish that nominal retinue distance was exhibited by the mean bee cluster for periods of time in both flight and GC modules. However, certain flight events caused observable disruptions to the retinue. Notably at t = 8.35 min, main chute deployment (Flight code: J) induced a shock force and increased queen distance in both S1 and S2, persisting for ~3–5 min, indicating that this event may be an important short-term stressor to queens. S2 showed increased queen distance for much of the flight time, yet the queen distance may have been impacted by the 3D beeswax topology of the bee compartment. From video footage and the retinue area measurement (Figure 2F, gray line), we observed that C1 worker bees wandered more generally and exhibited larger retinue area, while the flight worker bees (S1 and S2) clustered tightly. Tight grouping and lower mobility may be indicators of worker bees exhibiting stress and/or protective behavior toward the queen in the flight environment. However, differences in worker bee clustering behaviors may have also been impacted by the presence of beeswax structures. Post-flight, while egg size (a predictor of hive health) did exhibit a statistically significant difference in diameter from queens S1 and C1 (n = 6, CI 95%, p = p = 0.014), all observed eggs hatched into healthy offspring (Figure S3).

Overall, we delineate the merits of our generalizable module design: (i) The unified foam chassis provided positioning and securement for a plurality of module components and met structural and insulative spaceflight performance requirements; (ii) The design exhibited a weight-minimized chassis and thereby provided an unprecedented weight reallocation to experimental components, such as those commonly omitted due to miniaturization constraints; (iii) All components could be rapidly disassembled and reassembled for pre-flight checks and also for biological sample loading and retrieval. Finally, we documented original results from the autonomous experiment module and biological habitat for honey bees through take-off, microgravity, re-entry, and post-flight conditions.

Presently, it is a priority of public and private space organizations to expand the opportunities available to researchers to identify processes that are uniquely possible under microgravity (NASA, 2019). Moreover, biological experiments have been crucial to foundational research into long-term space operations—from investigating engineered food chains (Fischer and Laforsch, 2018) to addressing space radiation (Shunk et al., 2020) and human physiology in microgravity (Walls et al., 2020). Our experimental module was a first of its kind for insect space studies, and our platform provides an accessible format that can be easily adapted or augmented for future launches. Below, we discuss our contributions to general payload fabrication, video observation modules, and honey bee research.

Researchers are tasked with building custom payloads for experimental modules while adhering to performance requirements under weight and dimensional constraints. This high barrier to entry can be rethought to create generalizable design regimes using ubiquitous tools, inexpensive materials, and minimal parts. We created a design process that can be reproduced through a simple machining step with tools that are common to makerspaces. The CAD environment used to model the unified chassis simultaneously provided an intuitive way to layout design components and supported simulations of weight and thermal analysis before manufacturing (Figure 1C). Moreover, by using a press-fit assembly regime, the amount and complexity of housing hardware did not scale with the number of components. Thus, this method lowers the barriers of mechanical expertise and time, invites more diverse researchers, and provides the opportunity to test multiple iterations in the allotted development timeline.

Among the challenges of compacting a series of experimental capabilities, weight limits have heretofore curtailed the types and sensitivity of experimental equipment, including video quality (Cho et al., 2019). In the 2U experimental platform presented here, the design supports notably high image resolution (≤4K/3840 × 2160 pixels, equating to 0.05 mm resolution per pixel at the platform's focal distance), high frame rate (≤100 fps), and/or stabilization for video footage. This was achieved by modifying a compact consumer-grade camera solution (Hero5 Session, GoPro), using an analog switch to connect between the camera's solid-state shutter release and the onboard microcontroller (see Video S3 and SI Methods). Combined with emerging computational image analysis and machine learning processing, compact high-definition video modules are a powerful and versatile tool for space experimentation, and especially so for biological specimens.

At the level of honey bee research, while there are precedents for individual bumblebees (Yamashita et al., 2010) and a mature hive (Vandenberg et al., 1985) being brought to space, no prior missions have specifically studied the queen in relation to the effects of space transport. NASA mission STS-13 of 1984 reported on housing a hive in microgravity for 7 days and noted that eggs laid by the queen in the orbiter failed to hatch after the mission's return, raising many questions about bee reproductive viability (Vandenberg et al., 1985). By contrast, a queen in our study was able to lay viable eggs after spaceflight. However, due to limited subjects (n = 2) and flight duration, much is still unknown.

In addressing our experiment's limitations, we acknowledge that disparities between the GC and flight modules limited our ability to isolate causation of certain behaviors. The lack of beeswax in the C1 bee compartment created a less variable 3D space and a foreign surface material for the control bees, which could have impacted movement of bees as well as bees’ ability to thermoregulate or sense pheromone signals—as beeswax is a known medium for pheromone transduction (Yang et al., 2010). Yet, past literature on queen transport indicates that the absence of wax in short timeframes is not a major determining factor in long-term outcomes of queen quality (Withrow et al., 2019). We suggest that future experiments improve control setups, use pre- and post-flight recordings as datapoints, expose bees to analogue vibration environments in ground studies, and include additional sensors (e.g., CO₂ and volatile organic compound monitors, thermal imaging) to facilitate a richer analysis of the metabolic and pheromonal state of queen and retinue bees during flight.

Looking to the future, the platform presented here can help facilitate the expanded use of portable videographic recording and analyses techniques to track both retinue and queen movement in a broader array of studies to gain detailed information that is useful for determining both queen quality and the impacts of queen status on colony health in both traditional apiculture and space environments (Vergoz et al., 2009). Previous work on queen reproductive health in relation to transportation has focused on temperature as the major factor in queen health and subsequent productivity in the context of commercial pollination services and queen breeder shipments (Melicher et al., 2019; Withrow et al., 2019). In furtherance of the field, this study focused on differential gravitational, vibrational, and shock forces as sources of temporary retinue disruption, where temperature was held relatively constant. Thus, this research advances the technology readiness level for proposals of longer flight missions to further uncover insights into regulatory and reproductive functions in microgravity and whether bees may participate in our long-term space survival.

In conclusion, the generalizable method presented for constructing experimental modules provides broader accessibility to space research and new data on honey bee behavior in microgravity. This module design demonstrated a 25–50% chassis weight reduction compared to the state-of-the-art, and improvements in manufacturing time, simplicity, and cost that make space research more accessible to researchers. We further demonstrate that autonomous, video-enabled modules produced by this methodology can support and observe living specimens aboard suborbital flight and capture high-definition visual data on previously limited areas of space biology.