Artificial wombs are a staple of science fiction, but could we really build one? As time passes, we’re inching closer and closer to the day when it will finally become possible to grow a baby entirely outside the human body. Here’s what we’ll need to do to pull it off.

Top image by Mondolithic Studios.

More than just an incubator

A fully functional artificial uterus will be substantially more complex than a modern incubator, a clunky (and somewhat obtrusive) device that provides a preemie with oxygen, protection from cold, hydration and nutrition (via intravenous catheter or NG tube), and adequate levels of humidity.


Even in the best of cases, the current state-of-the-art doesn’t allow for viability outside of the womb until mid to late second trimester. Prior to that, a mother’s womb is the only option. Quite obviously, future incubators, or a full-blown artificial uterus, will push the limits of viability further and further until the entire gestational cycle can happen external to the body.

We’re still several decades away, but the two primary areas that need to be developed include biotechnology (for things like personalized genomics and tissue engineering) and nanotechnology (to facilitate micro-scale interactions and growth through artificial means). Smart computer systems and monitoring devices should also be developed to track the progress of the fetus’s growth, while automatically adjusting for changing conditions.

In terms of specifics, these are the broad components that will be required:

Artificial endometrium

The inner lining of the artificial uterus should resemble the real thing as much as possible.


Actually, for the first generation of artificial wombs, it would be prudent to mimic every gestational process as much as possible (we are producing a biological organism, after all). Later versions can then build upon what nature designed, and be optimized accordingly.

To that end, an artificial endometrium should not be made from glass or metal, but instead consist of a glandular layer made of real tissue. A blastocyst conceived via in vitro fertilization could then be implanted about 3 to 4 mm into the endometrium where it would take root and proceed to grow.


Work in this area has already been conducted by Cornell University’s Hung-Ching Liu. Many years ago, she prepared a co-culture system that combined epithelial and stromal cells (for ethical reasons these experiments weren’t extended beyond six days). Hung-Ching’s work is considered the first real attempt towards the development of an a-womb.

In addition to providing a physical starting point and enclosed space for the fetus, the artificial endometrium could also spawn and host a real placenta (e.g. by coaxing the growth of pluripotent stem cells), though it doesn’t necessarily have to come about this way.

Artificial placenta

And indeed, the growing fetus will also need a placenta, the organ which connects it to the uterine wall (via umbilicus) allowing for the delivery of nutrients, the elimination of waste, and gas exchange through the mother’s blood supply. Depending on the technologies available, the a-placenta could either develop “naturally” on the endometrial wall, or it could take the form an external device (or devices) that performs the same function. For example, a dialysis machine could actually help with waste disposal.


But a fully functional placenta will be crucial to the baby’s development and eventual good health. For example, the placenta is responsible for transferring the mother’s igG antibodies to the fetus — an important mechanism that provides protection to the infant while its immune system develops. Placental hormones also control fetal growth. During the early stages of pregnancy, the placenta provides the fetus with serotonin, which helps with brain development. And as noted, the placenta also regulates the way nutrients are delivered to the fetus, including the delivery of amino acids, fatty acids, and glucose.

The delivery of nutrients to the fetus should also reflect the way a mother would normally eat during the course of the pregnancy, both in terms of timing and composition of food.

If not designed and managed correctly, the fetus could experience problems, both in terms of growth restriction or overgrowth.


Getting an a-placenta to perform all these functions won’t be easy, but advances in personalized genetics and regenerative medicine will go a long way to make it happen. If our bodies can do it, so can a machine.

Fascinatingly, work on an artificial placenta has already begun. Goats have been kept alive for up to 237 hours in amniotic tanks through a process called extracorporeal membrane oxygenation (ECMO). It’s also a technique used in some neonatal wards to treat infants with medical problems involving gas exchange and the lungs.


Synthetic amniotic fluid

Dismissed as unimportant by biologists for many years, the fluid that fills the amniotic space is a complex and dynamic milieu. It changes as the pregnancy progresses (both in terms of its amount and composition) and it’s critical to fetal well-being. Producing and managing this ever-changing mixture will be just as critical as all the other gestational elements.


For example, amniotic fluid contains nutrients and growth factors that facilitate fetal growth. At first it consists of water and electrolytes, but it eventually contains proteins, carbohydrates, lipids, antimicrobial agents, and urea. It also protects and cushions the fetus. Image: Washington Times.

Amniotic fluid is also “inhaled” and “exhaled” by the fetus, an important process that’s essential to the development of healthy lungs. A fetus will also swallow the fluid, which creates the urea and meconium.

Temperature regulation

The incubator, if it can be called that, will also need to operate at just the right temperature. The fetus develops 0.3 to 0.5 degrees Celsius higher than mother’s, so typically about 37 degrees Celsius.


Proper stimulation

The fetus will also need to be stimulated across a number of sensorial domains. And indeed, the maternal womb has been called “an optimal, stimulating, interactive environment for human development.”

Ideally, the a-womb should move the unborn baby in a way reminiscent to how a mother moves, including standing, walking, and lying down positions. The incubator should be set to a 24-hour clock in which waking and sleeping hours are simulated. Basically, activity should never cease, nor should the fetus ever feel physically “isolated”. A sense of touch will also need to be simulated.


Fetuses are also active listeners. This is very important from a developmental perspective, both in terms of exciting the neural areas required for hearing, and for the unborn baby to bond with its caregivers in advance. Sounds should definitely be a part of the artificial uterus, including the steady swish-swishing of a heart beat.

Microbiome stimulation

It will also be important to kickstart a healthy gut microbiome. During vaginal birth, a baby is exposed to cocktail of microbes. This mixture ends up inside the baby’s gut where it helps them digest food, regulate bowels, develop their immune systems, and protect against infection.


To simulate this effect, biologists will have to recreate this mixture, ideally from biological samples derived from the mother (or grown externally). Image: Science Photo Library.

Final thoughts

An artificial womb will likely be the result of iterative attempts to push the limits of exosomatic viability. These days, the earliest that preemies can survive is around the 21 to 22 week mark. As time passes we can expect to see this number get smaller and smaller — and eventually to the point where a fetus can survive exclusively outside the womb. This will, of course, raise some thorny issues in the U.S. abortion debate.


Once in use, and after some time has passed, scientists will undoubtedly study the long term effects on babies born in an a-womb as compared to those born naturally. Initially, the health of a-womb spawned babies will likely be inferior to those grown in a real womb. Refinements will undoubtedly improve these results over time. And in fact, I wholly expect that an artificial womb will someday bring a baby to term in a way that’s even superior to the real thing.

Indeed, unlike a mother, an artificial womb is not susceptible to disease or malnourishment, nor will it be prone to drinking or smoking. And with the assistance of powerful computers, advanced biotech, and even microscopic machines, the gestational process will be further optimized.


It’s also interesting to consider how this technology will be received, and how many people will opt into it. It’s safe to say that many prospective parents will react negatively to it, arguing that natural will always be best. But for those who need or want it, the advent of artificial wombs will certainly herald an unprecedented stage in human history.

Additional reporting by Joseph Bennington-Castro.

Source not cited: The Artificial Womb (2011).