This is an entry for the Growing Beyond Earth Contest in the "professional" category. And by professional I mean I've grown a few container gardens and own some Arduino UNOs. I wasn't going to enter the contest. I mean this is NASA after all, there was nothing I could possibly contribute, right? But I subscribe to a lot of the NASA news feeds and they are always talking about experiments (chocolate chip cookies, oh yeah) so I kept turning the problem over in my head. It was actually a fascinating problem and attempting to improve the existing growing system challenged me and kept me happily engaged for hours.

Then I realized something, sharing information and dreams is a big part of what NASA is about. So my solutions might be simple, but participating in that dream, contributing to that challenge with optimism is worthwhile endeavor in and of itself. I'm an adult and I'm still dreaming because of what NASA is doing, so I kind of had to contribute something.

Like most design problems, I both overcomplicated and oversimplified the problem before finding some design ideas that might actually be useful. The specifications for the design are simple:

Adhere to the volumetric constraints (a 50cm cube)
Contain all necessary features for plant growth (light, water, etc.)
Use 3-dimensional space in an inventive and effective manner

Most of my design ideas revolve around the "Use 3-dimensional space in an inventive and effective manner" criteria. Looking through the NASA documents started to give me some ideas. The photos showed lots of wasted space around the plants. NASA is aware of that inefficiency and it is the focus of the this contest.

At first I thought, "Well just use succession planting, selective harvesting along with square foot gardening techniques." I have grown several compact gardens, so thinking about the plants' foliage and root habits, growing periods and how different combinations can be intermingled in a small space is a common practice for me. All these techniques do help, but there would still be a lot of empty space.

Then I realized that plants on the ISS could face any direction and use multiple light sources. That's when I started thinking of the garden as a 3d packing/nesting puzzle. But things got even more interesting because of how water and gases behave in low gravity, the need to contain water/plant/soil/nutrient in a single package, and the need to use the same items multiple times using limited cleaning methods. After that, I came up with a lot of ideas, some simple, some nifty, and some kind of crazy.

Step 1: Frame and Supports

"If NASA is looking for something new I must abandon all previous solutions and create something that's never been seen before," or so my early thinking went. I did thought experiments and doodled inflatable origami containers, ornate 3d printed honeycombs; and I even considered growing bamboo into forms to create lightweight structural members. I like all these ideas, but they aren't the most practical for a space station.

Good Old 2020: After thinking for a while, and reviewing the NASA docs, I decided to stay with standard 2020 extrusions. Yes, that is boring. But the extrusions are dependable and they can be reused, reconfigured and replaced easily. Numerous attachments are commercially available so corner braces, rails and panels are already available. A custom plastic bubble or flexy honeycomb have few, if any other uses on the ISS.

My frame design keeps most of the extrusion exposed for easy access and future use. The current grow chamber has much of the extrusion wrapped in what appears to be sheet metal. Any sharp edges or corners on the extrusions can be covered by standard end caps, making them almost as safe as the metal wrapped versions.

I was especially concerned with keeping the ability to easily reconfigure the frame and attachments. Many of my other design ideas require changing the location, quantity and orientation of the lighting panels, grow platforms and support systems.

That Seems Too Simple: At first I designed heavy and complicated structures, sturdy enough to support plants and water in mid air here on earth. I added some motorized movements and sliding mechanisms. But then I realized the plants and containers on the ISS almost float because of the low gravity. And with a movement range of less than 50cm, it seemed overkill to add stepper motors with screw drives or a scissor lift. I wasn't designing a CNC. All I needed to do was keep the plants and panels floating in place, not really support them against the pull of gravity.

Positioning and Securing: Off-the-shelf corner braces and connectors can be used. A DIN Rail or other easy to adjust sliding system with a cushioned spring-loaded clamp at the end of a swivel joint or "flexible" gooseneck arm may be all that is needed to position the lights and grow platforms. Any specialty attachments or modification should be simple enough to be manufactured on board the ISS with magnets, Velcro, InstaMorph, Fun-Tak or similar.

Clips will be required to secure and position the lighting panels and wiring, the growing platform, the water lines, the sensor-data/power lines and the fans along with their power lines. A small ferrous metal rail may be needed to attach the clear canopy. At two of the corner edges, a simple string or cable might be enough to hold objects in place while keeping maximum access.

However, except for that one terrifying moment on a swing set, I've never actually experienced weightlessness. So I might be oversimplifying things.

Peripherals Panels: Two of the "walls" of the grow cube are assigned to housing support systems. I wanted to keep the grow cube as open as possible, both for ease of access and viewing pleasure. But if needed, additional sides can be occupied my metal panels to facilitate magnetic or physical mounting of peripherals.

One wall panel is assigned to electronics and power distribution. All sensors and actuators feed back to this panel. it is also the main interface to the EXPRESS rack system.

The second wall panel houses the water, nutrient packs, valves and extra tubing. Displays and controls for these systems are located on the exposed edge of the panel.

Connection points to the ISS power, data and other systems will be placed as needed.

Curtains: Because the configuration of objects changes over time in most of my designs, the current clear pulldown sleeve/wrapper might not work. Instead, sheets of clear plastic with a few magnets might be enough to allow easy access and viewing while maintaining a closed growing environment. A way to open and close ventilation ports in the canopy may be required because of the addition of more "growing levels" and the ability to change fan positions. The fans might need to placed anywhere on any face, then moved later in the sequence. However, if the fans' intakes and outlets can be ducted, a standard opening location might suffice. At the volume of air we're talking about, a piece of bent paper might be all the ducting required.

LATER: As experiments continue, a dependable sequence of plant growth will be created. The all-options-all-the-time system can then be standardized down to a few options with more standardized mounting and attachment defined.

Step 2: Sensors, Actuators and Electronics

The electronics for control and data are to be housed in a metal panel case on one face of the grow cube. Connections for external power and data should be provided. Routings for connecting to remote pumps, fans, sensors and other actuators shall be provided. It is anticipated that the electronics and sensors will consist of multiple, commercially available components (see below for details) instead of an all-in-one custom system.

Although it isn't the primary focus of this contest, I would like to offer some ideas for consideration.

Create guidelines for a system based on commonly available hardware and software. Specifically, a system based around the Python programming language. This would allow students and tinkerers to participate on a variety of hardware. Python is growing in popularity and works interchangeably on a variety of platforms, from Adafruit's line of Feather microprocessors to the M4 Grand Central board from the same manufacturer, on a Raspberry Pi or a full computer system. Their is already a precedent for Python in space.

If something like the STEMMA connection system was used, students and makers could acquire and reconfigure a variety of sensors and actuators that share a common connection system. With the STEMMA system, experimenters can reconfigure sensors without soldering and with a minimal use of IO pins. There are already a wide variety of sensors and actuators available, and turning an existing sensor breakout board into a STEMMA compatible device is easy. These hobbyist level sensors may not be acceptable for use on the ISS, but gathering data with them, interpreting the results and taking actions based on the analysis of the data should be similar enough for prototyping and education.

I can see the need for several sensors and actuators (Most available directly from Adafruit):

Water sensors
Humidity sensors
Temperature sensors
Dissolved Oxygen sensor
Pump(s)
Solenoid water valves
Fans
Grow Light Addressable LEDs
Indicator LEDs
Heat/Cool Source
Graphic Display
Switches
Buttons
Rotary Encoders
Speaker/Piezo

Such a system would allow students and makers to work with familiar systems and learn to connect their power and data systems to other standards. A common platform would allow sharing and coordinating of development for secondary challenges such as menu systems using more sophisticated displays, creating automated control and feedback systems, analyzing data and communicating with other devices.

Whether and how long data should be stored in the grow cube is an unknown for me. What kind of analysis the cube is expected to perform is also an unknown. These requirements will determine what kind of processing power and memory/storage space is needed.

As a next step here on earth, real time data sharing of student experiments (or even from the ISS) could be done on the Adafruit IO platform. Or the data could be sent to a central Google Sheets repository. If multiple teams share the same basic planting scheme, it would be easy to compare the effect of minor changes to the watering schedule, lighting, nutrient injections etc.

Remote monitoring of ISS sensor data and imagery by citizen scientists could offload some of the repetitive tasks from the station crew. One garden cube might seem a luxury, so monitoring it is a pleasure. Monitoring 12 or 24 cubes with different configurations can become a chore. Sharing the tasks with other teams and creating ownership of a particular garden might prevent any one person or group from getting overwhelmed. Plus I know I would really enjoy knowing the crew just ate a salad from "my" garden.

Step 3: Water, Light, Nutrients

My first interpretation of the challenge made me think the design had to carry enough water, gases and nutrients inside the 50cm cube for the entire growth cycle. However, after more reading it seems that periodic replenishment is allowed.

WATER:

Enough water for about a week can be stored in the flat container located in the "top" panel of the frame. A single peristaltic pump and food safe silicone tubes feed the water to the plant pillows. A series of manual or automatic valves control the amount of water sent to each plant pillow (or plant type). Injections of fertilizer dissolved in water can also be dispensed to individual plants or plant types to meet their differing needs.

I keep reading about oxygenation of the water and flooding reactions. That's easy to solve on earth, percolation provides oxygen and something called drainage is easy to achieve with gravity. How to oxygenate water in space and how to control and direct the movement of water around the plant roots without the aid of gravity is still something of a mystery to me

NUTRIENTS: Currently, nutrients are dispensed via time-released fertilizer incorporated into each grow pillow. I thought about putting a nutrient injection option near the water pump. This would be good if only one plant type of one age was being grown. But if multiple plants of multiple generations are being gown together, a more customizable supplementation system might be required. A simple addition might be an access point for injecting liquified nutrient supplements near each pillow.

LIGHT: Many of my designs require increasing the number of light panels. Some designs require two while others could require four or more. Some of the more advanced designs may require smaller snap-together panels or even flexible panels, "sticks" or strips. All types of light sources should be easy to reconfigure and hold with the clips available on the main frame.

Since most LEDs are highly directional, a series of strips or tubes might be the most versatile form. They can be mounted to the side or rear panels to mimic the current configuration. But they can also be clustered into a central radiating light source. They can be used one-at-a-time with seedlings, ganged into pairs or trios, and if magnetically mounted to the frame, can be quickly moved to provide the optimal lighting for each planting area's constantly changing needs.

Moving the lights might be simpler than moving the plants and growing pillows.

Adding "addressable" LEDs to the light panels would help keep illumination even for individual plants. Closer plants might not need the lights at full strength, but smaller plants or those farther away from the LEDs might need increased light levels to get an equal amount of energy. It might also help with future phototropic foliage training experiments (topiary with photons) by allowing the crew to illuminate specific areas to train plants to grow towards that area and increase the efficiency and crop yield.

Opinions vary, but "daylight balanced" (5000K-6500K) LED panels may be sufficient to grow plants in the proposed environment. Panels and strips of these lights are available in a variety of sizes. Some are on a flexible backing which would allow even more options for positioning. Many are "addressable" so individual LEDS can be turned on or off and dimmed as needed. Adafruit sells a variety of such products.

AIR and CO2: Because most of the designs have multiple levels and orientations, it is critical that the fan (or fans) can be easily moved and their output directed as needed. The fan itself, and possibly some ancillary duct boxes (chases), can be attached directly to the frame extrusions to provide air movement for gaseous, heat/cool, and humidity control. I am unclear on whether additional CO2 is required. And I never saw this mentioned in the literature, but multiple fans might also aid in plant movement and facilitate capillary action.

Step 4: Plant and Root Shapes, Mix Them Up

Plant foliage shapes and growth patterns vary. Their rooting habits vary, and their light and water requirements vary. Gardeners who practice succession planting and square foot gardening methods use these variations in shapes and growth habits to their advantage. Arranging the location of plants and coordinating the timing of plant-growth helps maximize yield over the course of a growing season.

Multiple crop types can be grown in the same area (volume) as one crop type. Mixing different foliage characteristics in overlapping patterns and growing multiple generations of plants of differing heights increases food yield per unit volume.

Maximizing an array of a single type of plant for a given volume might be more difficult than maximizing a mixed collection of plant types. Plants naturally adapt their shape to conditions, and they can be artificially constrained, trained and pruned to grow in shapes slightly different than their normal habit. But there are limits to the changes that can be made to a plant's natural shape before you start loosing productivity.

Small plants can fill in space left over by larger plants. A lacey umbral can grow over the top of other plants with minimal impact. A ground-runner can flourish under the shade of other plants. Green onions and shallots can be sprinkled like pixie dust and will poke their tasty, slender shoots through the other plant's foliage. A carrot or radishes can form underground while the next generation of greens grow on the surface beneath the carrot's leaves.

This intermingled, intertwined overlapping way of gardening is just a more ordered way copying what nature does. One of the most productive plots I've ever grown resulted from a rainstorm washing a random collection of seeds into a gully. I was bummed that my perfect garden design got "ruined" by nature, but that gully turned out to be a mixed up tangle of every plant type in the garden. As much food came from that tangled mess as from my well designed and carefully tended garden.

I am not advocating throwing a random collection of seeds into a grow pillow. But I do believe that plants can thrive in a much less manicured, less grid like structure than many gardeners believe. More experimentation with this technique is warranted, and most of my designs use this concept..

All of this is relatively simple on earth. But the ISS uses grow pillows; you cannot just plant things at random and let their roots and leaves intermingle. Because the ISS space garden uses pillows of a specific size instead of a single "open plan" plot of soil, creating the arrangements and sequence is more complex and would require stacking, rearranging or exchanging the pillows to create similar sequences and results.

Luckily, plants are incredibly adaptable. I have seen foliage reach around obstacles to get to light or completely change orientation mid-growth when their container falls over. I've seen hemispherical and tap-root system adapt from their natural shapes to become flat, planar, linear or maze-like networks to get around rocks and other obstacles. Farmers intentional grow pears inside glass bottles, make melons cubical and grow tomatoes upside down.

If the shapes, sizes and locations of the grow pillows are arranged and sequenced correctly, and if the phototropic nature of plants holds true in space, you might be able to stack and overlap the grow pillows to create a varied mix of plant types that takes advantage of all the space available inside the 50cm3 growing volume.

Step 5: New Soil Pillow Shapes, Sizes and Openings

Many of my design patterns show small plants grouped tightly together. This isn't possible with a full sized grow pillow. But if the plants were germinated in a mini-pillow, many more seedling could be started in a smaller space and the strongest seedlings could then get transferred to the larger pillows. If the mini-pillow, basically a small plug of growing gel (pullulan) had a frame that snap-fit onto the larger pillow, the gardeners could easily move individual plants into pillows that are in the most efficient location for the current garden arrangement.

Currently, it appears that all grow pillows are about the same size and shape. However, my experience indicates that roots and foliage can adapt to and prosper in almost any reasonable shape of a container or growing volume. They can also share territory and grow happily with their roots intermingled and leaves intertwined as long as their basic water and nutrient requirements are similar and they are not antagonistic towards each other. If this holds true in low-gravity, then the options for interplanting and sequencing environments is greatly increased.

With this in mind, I propose experimenting with both differently shaped pillows and pillows with multiple "ports" for growing plants. A handful of shapes and port arrangements would be enough to test the idea.

Grids and Order: A grid system based on 3cm increments seems reasonable. Soil depths of 3-6cm may accommodate most plants. Width and lengths of 9-12cm may be good for many plants. Larger pillows may be need for multi-plant arrangements. And deeper "tower" shapes can handle tubers and root vegetables as well as tuck into otherwise empty corners and make them productive. A handful of standard sizes can handle most plants appropriate for the designated growing volume.

By combining the standard sizes, developing a layout that totals 42, 45 or even 48cm is possible. That makes use of most of the 50 cubic centimeters.

New Shapes: Use the shapes of pillows and their arrangement to elevate plants towards the light or recess them to make room for other plants. Use new shapes to accommodate root vegetables. Carrots, radishes and other root/tuber vegetables may need a much deeper pillow. The same holds true for plants with tap-root systems. Chives and green onions may only need a shallow pillow.

A triangular pillow might work better for a perimeter/center arrangement. Triangles would fit the frame's corners if the plants were on the perimeter growing towards a central light. The same triangle shape could be clustered into a center growing-column with lights placed around the perimeter.

Multi Plant Pillows: Growing the next generation of plants in the shade of the mature plants is a common practice. Using multiple growing ports in the same pillow would allow a succession of the same plant type to be grown in one pillow.

Or use the multiple ports to grow several types of plants in the same pillow. Establish a primary plant for each pillow, but have additional ports for smaller plants. If lettuce is the primary plant it would get planted in the central main port, but green onions and herbs could be planted in the secondary ports.

When combined with the ability to point plants in any direction, the 3d options only increase. A root vegetable can have its roots occupying the empty corner of the greens crop growing in the opposite direction. A slender crop could be growing 90 degrees out of phase spatially from two other crops growing 180 opposite each other.

Is This Too Many Variables: I realize that quantifying plant needs and performance is already difficult with one pillow shape and one growing port per pillow. And I have debated the benefits of mixed plantings versus traditional gardens with several row-and-grid gardeners and never converted any of them. But I have won grudging admission that small plots of "messy, cottage garden" style interplanting can yield as much food as larger tilled and tended grids. So if I had to change my gardening style to work on a four foot apartment balcony, the same techniques might be worth exploring on the ISS.

Step 6: Patterns for Maximum Growth in 3D Space

When you put the previously outlined systems together, you get a classic packing challenge. Differently shaped and sized pillows need to get arranged in a system that gives all plants what they need while using as much of the growing cube's volume as possible. The arrangement might change as plants mature and are harvested.

At first glance, this might seem like an overly complicated system. However, I have done similar in earthbound gardens and once the sequence is designed, there is little extra work required. True, for the first year I had a spreadsheet of foliage types and growth times, plus a Gantt chart to help me plan the sequencing and placement. But after the first year, with the knowledge stored in my brain, the planning and sequencing was second nature.

Keeping the pillow system and adding these changes does have advantages over changing to a "hard coded" hydroponic pipe system. The gardeners are freer to experiment with combinations, and the ability to move plants to new locations with varying spacing allows them to constantly maximize the space. With more rigid systems, it is more difficult to maintain the maximum usage across the entire growth cycle. I don't have the math or animations to prove this claim. But my experience shows that I can grow more food in a smaller space with these techniques. I've also done the traditionally rigid, row and grid gardening and the yield per unit space is much lower.

There is a wide range of patterns to work with, and balancing optimal usage of space against labor and planning will determine which pattern and sequence is chosen. A spreadsheet might be needed. Oh wait, this is NASA. They probably already have a good packing algorithm.

Does It Scale: I was worried about transitioning this system to a large scale, robotics based system. Then I realized that I still think of robots as simple things. I might still be struggling to control a robot on a simple XYZ rail system, but very agile robots are common now. And us hobbyists can already do object detection with "edge" boards and a cheap camera. So having some green onions poking up through the spinach might not be as confusing to robotic farmers as I had originally thought.

Work The Problem: I began to design different growing patterns. I tried to hard-code the arrangement of the grow pillows and wound up just as restricted as I did with a hydroponic pipe. And I could never find the perfect arrangement for all possible scenarios. So I started thinking of the plant pillows and arrangements more like a kit of parts, a LEGO kit for space gardening. I can't possibly anticipate every configuration. But I can advocate for a freer way of thinking about how the pillows could be arranged. Maybe, I decided, a series of schematic patterns and some cheerleading would be my best option.

I started with the simplest change, have two growing platforms, each growing 180 degrees in opposition. This could be splitting the growing chamber horizontally, vertically or even diagonally. The division does not need to be symmetrical. A staged movement of the growth platforms could allow multiple generations of crops to be growing at the same time. A large division could be supporting nearly mature plants while a smaller division could support a seedling bed for the next generation. The number of divisions could be increased to one each for seedlings, young plants, established plants and plants ready for harvest.

Then I ran the same exercise on center-loaded plantings and edge-loaded plantings. The golden ratio cube method might also work. Even with all these options, there were still empty spaces around the plants. But with a stick/tube light and some Velcro, it might be easy to put a plant in the corner temporarily, then move it to a new location for maturing. My head was swimming with possibilities and I had a grid of patterns to choose from.

Just Start: I have been in this situation before, my head swimming with possibilities, and I think the same solution I used then will work here. Pick five plants and a low number of generations and design a system for that. It hurts to give up those thousands of other possible combinations, but do it. Design a sequence and planting pattern for the chosen five plants. grow the plants. Take notes about how it works. Use those notes to improve next year's garden design. As an added benefit, something will change the pattern for you mid-cycle and a space for additional experimentation will open up. Bugs, deer, rabbits, weather, or a toddler are all likely agents of change on earth. Odd behaviors of liquids, gases and soluble nutrients are likely culprits on the ISS.

Basic Layouts: Different sizes and shapes of growth pillows might allow

Simple split - horizontal/vertical/diagonal
Reversing checkerboard
Multi-level reversing checkerboard
Troughs
Planes and towers
Center Loaded
Edge Loaded
Suspended sphere with encircling foliage

Step 7: Soil Pillow Alternatives

I know improving the soil pillow system is not the primary focus of the contest. But the NASA scientists did encourage participants to think about other aspects of the system. After designing many ways to change the shapes and arrangement of the current soil pillows, my mind often returned to finding a more adaptable alternative.

Deployable mechanisms seemed like a promising avenue. So I doodled some:

Auxetic meshes
Hoberman sphere
Overloaded flexible sphere

The grow-pillows are expensive to ship and can only be used once. So I tried to think of something that could be reused multiple times, could be cleaned easily, and could be reconfigured for different plants and planting arrangements. A method to control the water and oxygenation of the roots was also a concern.

Silicone seemed like a good material. It can be food and skin safe. It is chemical resistant so it can be cleaned a reused. It can be very flexible or fairly stiff, and mixing parts made from different hardnesses can create parts with unique capabilities. Couplings that bend but grip their counterparts snugly enough to be watertight are possible.

Silicone also has a "shape memory" and wants to return to its original shape after deformation. This allows the creation of chambers that can be squeezed flat, but that return to their original shape when pressure is removed. If check valves (or flaps) are added, the chamber can act as a pump.

Silicone is also easy to repair, and new parts can be cast for replacements or iterative improvements.

The water supply system needs to inject water and extract it to control the level of water, nutrients and oxygen. Mechanical pumps can move water and air. But to control the quantities for twenty or more plants would require multiple pumps or a complex branching system of inputs and outputs. Maybe a more organic, biomimetic approach would work.

I began doodling designs modeled after soft robotic actuators. I have made several types of silicone actuators based on XYX Aidens' (and others) wonderful Instructables. I know they are strong, flexible and can be fine tuned in place in a variety of ways. They can be flat, linear, spherical or coil around an object. Their movements can be sequenced to achieve sophisticated types of movement. And they are gentle enough to use around delicate plants and their root systems.

The far reaches of wild ideas had an octopus-like arrangement with tentacles featuring plants growing in place of the suckers. The large bulbous body (head) acted as a large, squeezable pump system. Bioluminescent effects were desired but not designed.

When I toned down the design I was left with something between "milking a cow" and a "beating heart for plants." It is essentially a traditional hydroponics system but with flexible veins and pump/grow nodes instead of a hard, inflexible tube. The flexibility allows garden designers to sequence the arrangement of plants and vary the arrangement over the course of the crop cycle.

If arranged as an auxetic or Hoberman mechanism the plants could sprout in a tight space and, as the plants grew, the growth chambers could be pulled apart or bent into arcs to make more room from the lush foliage of the plants. It consists of a silicone sleeve that fits over an opaque PVC (or maybe Pyrex or acrylic) hard chamber for the roots. Filters between the silicone pump sleeve and the hard root chamber allow water to move but prevent incursion by roots (think pumice stone filter on a household spigot. Although mechanical pumps provide the majority of water movement, maybe a "bulb and flap" soft pump could be added to the silicone sleeve. A squeeze from a human or robotic hand can add or remove water to tailor the level for individual plants. If the pump connected to several tubes and chambers, a type of peristalsis might be possible.

But that was still overly ambitious, so I settled on a more basic system of hard grow chambers fed by a common water line and emptying into a common extraction line. It's not really auxetic and not very organic feeling. But it does offer the ability to move plants mid-growth just by bending the silicone lines. And the ability to configure the lines and spacing of the T-feed couplings allow the designer to adapt the design to fit almost any space. By combining multiple grow chambers for similar plants or predetermined planting patterns into a single fisture, the system can easily be scaled up.

Step 8: Conclusion

It feels odd not to have a working prototype, no code or step-by-step instructions, or even a fancy CAD rendering. But I have grown significantly more than 12 plants in a small volume while being constrained by gravity. So I know the packing, interplanting and succession techniques work. Going full 3d could be freeing, but also present challenges. Any of the design ideas presented here would be a fascinating project to pursue, even if I'm using potting soil and cardboard to mimic the ISS system.

And yes, I realize the primary focus of this contest is growing more lettuce. But I kept designing for herbs and allium and crunchy things because I have also lived in a cubical and coming home to play with my balcony garden, run my hands over some basil just for the scent, chew on chive and chill was a highlight of my day. The NASA docs refer to it as "organoleptic" concerns. I call it enjoying nature, or even more simply, gardening. And this 50 cubic centimeters is still a garden, as full of nurturing potential as any other.

Thanks for sharing the dream.