Sunday, March 13, 2016

Saga of the space turtles: Intergalactic transmission of life using today's technology

Spaceship  from Star Trek: How close is science fiction to reality ?

The fury of an idiotic race: 
Out, out, brief candle!
Life is but a walking shadow, a poor player 
That struts and frets his hour upon the stage
And then is heard no more. It is a tale
Told by an idiot, full of sound and fury
Signifying nothing.
This quote is from Shakespeare's Macbeth. It signifies that human life is like a soap bubble: a minor shimmer of existence before it bursts and disappears into death. And such is the fate for the entire human race.  

We humans are living on this planet for an extremely tiny glimpse of time. We may vanish altogether in an instant. All of our cousin species in the Homo genus went extinct. Some of them have indeed lived for much longer than us. For example, Homo Habilis lived for 1-2 million years on this planet.  In contrast, the history of Homo Sapiens is less than 0.2 million years old. Even within our species timeline, the vast majority of our history is unbeknownst to us. We spread out of Africa around 60 thousand years ago. What follows is a ridiculously small amount of time. In comparison, the light from the center of our own galaxy - the Milky Way - will take 26 thousand years to reach us. The Milky Way is around 100 thousand light years across. The nearest galaxy Andromeda is 2.5 million light years away.

It is only about 60 thousand years ago that we started making stone tools and producing cave art. The history is human culture is far shorter. The earliest written alphabet is just 5 thousand years ago. The great religions that are still alive in our society arose even later.  Thus, all the bickering and fighting we do between nations and religions is for a nanoscopic length of evolutionary time. In physics, at scales as minute as this, quantum effects predominate. Random particles and their antiparticles will arise and die in the very next instant. The history of the human race - all our hopes and dreams, passions and petty fights - they all compress into such a meaningless trivium. We arise and disappear before another intelligent life in this universe, if it even exists,  bothers to take a look at the earth.

The physicist Carlo Rovelli says

We are not a species that is known for sticking around. We are not like the turtles, which have lived unchanged for millions of years. We are very much like the other Homo species.
Think like a turtle: A Space Turtle

Our lives on this planet may be short. But before we disappear, we will be tearing down an entire fabric of life. Since our very origins, we humans have been systematically destroying ecosystems and scorching this planet. When we first spread to Australia, about 40 thousand years ago, the continent was green and teeming with life. Soon after humans arrived, grand and gorgeous species of mega-fauna, that lived for millions of years on the continent, went extinct: either hunted down by us, or eliminated by the disruption of the ecological balance by us. The anthropologist Jared Diamond explains that this destruction of ecosystems had terrible consequences for the future development of humans in Australia. With noting to sustain large human societies, the populations began to perish and those who survived had to revert to a primitive lifestyle with minimal technology. Some human populations in Tasmania have even forgotten how to light a fire. Diamond argues that the advantages of civilization are only temporary, and that mismanagement of ecological resources will spell death to civilizational progress. There may be a short explosive burst of death such as wrought by a nuclear war, or a slow painful drawl of death due to famine and pollution. But the end result is the same.

Whatever be our fate, we modern humans bear an intense fury of destruction that would shock even our ancestors. We are turning oceans into acid and the atmosphere into a grand sewage dump for carbon dioxide. As the planet warms due to our actions, we await famines, extreme weather and scarcity of fresh water. We do not know how much of technology, ecological resources and civlization we will be leaving in tact for our future generations. There may not be any large species of animals left on this planet, except those that we imprison in concentration camps called as industrial farms. All big species of animals and plants may go extinct in the wild.

If the rate of species extinction is awful on land, it is even more terrible in the oceans. We humans have been relatively late to make our presence there, but we are making up for the lost time. Many whales, sharks and fish are now facing extinction. Majestic animals like the leatherback turtle, which crosses the entire length of the Pacific ocean in a year, subsisting on nothing but jellyfish, may vanish from earth.  These gentle creatures have survived for millions of years in the same lifestyle. One of my favorite authors - Caspar Henderson, in his "Book of barely imagined beings", recounts a magical moment touching the back of a leatherback turtle as she was laying eggs.
Touching the leatherback holds all the magic of childhood. The animal is vividly alive in a realm that is largely beyond our reach and our imaginations. Later I recall a phrase attributed to Zhuangzi: 'all the creatures in this world have dimensions that cannot be calculated'.
Often, the destruction of life by humans cannot be seen but can be heard by the ears. The musicologist and naturalist Bernie Krause talks about the spectrum of biosymphonies, where different species find and occupy their own niches of frequency, much as individual musical instruments find niches in the timbre of an orchestra. When an ecosystem is damaged, its sound signature changes drastically. We humans are thus systematically destroying the music of this planet. Beautiful sonorific places teeming with life, such as the Amazon basin or the Congo delta, are being turned into silent graveyards by us, disturbed only by our anthrophonic sounds of the motors of vehicles and machine saws.

But is there any hope to save this spectacle of life on earth ? Is there any hope for us - the idiotic race of this planet ? In the grand vista of this universe, there may be no other place as vivid with memories and beauty as our own planet. Can we preserve these memories and relive them in a different time, and in a different planet ?

This is not easy. There is no other planet in our solar system or in adjacent star systems which is even remotely like earth. So this magical spectacle of life cannot be recreated nearby, but how about traveling into deep space ?

Saga of space turtles:


Pale blue dot: where is land ?
 The Pacific ocean on earth is a microcosm for deep space

The first thing we should note about deep space is that it is vast. There is nowhere for us to alight and take a short break. Barring a grand revolution in physics, possibly resolving the mysteries of quantum entanglement and the statistics of dark particles in vacuum, intergalactic travel will need millions of years. There will be no pit stops for us to refuel our tanks. A comparable analogy is humans trying to cross the Pacific ocean. Let us first consider the problem of exploring deep space. Because we do not even know where to go.

In 2014, NASA's Kepler mission found an 'earth-like' planet in the habitable zone. This planet, known as Kepler-186f is 500 light years away from us. This planet may very well be a red flag, but reaching even here will take us at least millions of years. In reality, what we need is a program of constant space exploration to hunt for the right planet. This exploration needs to run for hundreds of thousands of years, spanning across millions of watch-towers in space. We cannot wait for the perfection of science and technology to start this program of space exploration. In fact, we do not have any guarantee that our earth will remain intact for much time longer (probably due to our own foolishness). So we need to start working now: we need to build a space program that can be updated as technology advances in the future. Referring to the Viking sagas and the leatherback turtles, I will call this space program as the saga of the space turtles.

What do I mean by a space turtle ? It is a small self-contained space ship that can be built in a modular fashion. It would have protective outer covering for dealing with cosmic radiation and dust projectiles in space. Inside, it will have long-lasting electronics and robotics gear. It will be equipped with radio and laser transmission equipment, as well as spectrometers and other instruments to do  astronomy.  The optical equipment inside the space turtle should be made such that they have no single point of failure. This can be achieved through recent advances in computational photography, with millions of sensor cells and thin optical components. Similar to the eyes in humans and animals, these computational eyes will be robust to wear and tear, and will rely on computational intelligence for doing most of the interpretation. Thus, each space turtle will be a hardy robotic animal, that can be mass-produced and sent to explore deep space.

First, we need to build a factory outside the earth's atmosphere for fabricating these space turtles and launching them outwards. Capitalizing on recent advances in 3D printing and electronics, we can build such a factory in a relatively small space, possibly by revamping the international space station. It is necessary to do so because sending a new space-turtle should be a relatively simple procedure, which should not require huge budgets simply to launch it beyond the earth's atmosphere. Sure, it is still necessary to overcome the earth's gravitational pull, but it is possible to do this using  the alignment of the moon and other planets and minimal fuel otherwise. Thus, it will become cost-effective to make far more launches than is currently possible.  This is important because, when we are exploring a brand new territory, the ability to perform many experiments increases greatly our likelihood of success. But equally important is our ability to learn from mistakes and update our knowledge. With respect to the space-turtles, it means we need to be able to update their operative procedures from across space: what is their navigational orbit, where are their lenses focused, what type of image analysis software do they have on board to analyze the astronomical observations etc. We should be able to do this as simply as performing a software upgrade across space.


Performing a software update in space:

The Opportunity rover on Mars: A great achievement for software

In any space mission today, the on board software is sealed and finalized several years before the launch. I once visited the NASA Jet Propulsion Lab and met with the engineers (my heroes) who sent the Spirit and Opportunity rovers to Mars. These rovers were remarkable achievements of precision in landing and navigation, but the software they had on board was at least a decade old. I was completely dumbstruck when I was told this. In extremely fast-moving fields like computer vision and robotics, this means this software was from a completely different generation of algorithms. This makes it all the more remarkable how it worked perfectly and we were able to achieve such results in navigation and landing.

Space exploration is extremely costly and there is very little room for failure, so a battery of tests are done before the launch to check that the software works under every possible criterion. In fact, rigorous theorem proving and software verification are done to their highest standards in the software meant for space missions. However, these practices are exactly the opposite of how we build large software systems today on the internet. Here, we perform software update at the speed of light and even connect the computers and electronic devices perennially to the internet. We access not only data but also programs through cloud computing. An army of software developers and freelancers collaborate on adding new features and see their results instantaneously. This speed and flexibility in development enables us to take risks and test millions of possible options. It is currently not possible to do this for space missions. But why not ?

A space turtle launched from outside the earth's atmosphere will be a lower risk mission than sending it from the earth's surface using rockets. Secondly, the "intelligence" required in a space turtle will be limited to astronomical navigation and observation, at least initially. This will be much easier to accomplish than greater challenges such as soft landing, or path planning on an exoplanet. So in principle, these software updates can be sent by radio transmission or by optical laser transmission, if there is a direct line of sight to the space turtle. The data that can be transmitted will be then limited in bit-rate, but we are thinking in terms of millions of years to explore deep space. So this is not a problem. When we can perform software update in space, it will be possible to transmit large objects such as the latest neural networks to recognize specific patterns on astronomical images, or those to  optimally navigate and orient the instruments aboard the space ship.

Building an intergalactic network: 

Path of the Voyager mission: The first man-made object to escape from the solar system

When the internet was built in the 1970s, the engineers used to call it the intergalactic network. This was a vision and inspiration for them to build it in a completely foolproof manner, able to withstand large shutdowns of relay servers or even a nuclear attack. Those engineered systems from the 1970s have been a huge success. The internet is an extremely reliable system - probably, the best that the human ingenuity has produced so far. 

So, when do we build a real inter-galactic network ?  Using space-turtles as relay stations, they can be made to communicate with each other and evolve a protocol for transmitting information across space. In order to do this, they have to be dispersed widely to  various locations throughout the solar system, and later, to outside into the deep space. A space-turtle will not be tethered to a fixed geographic location, such as an internet relay point on earth. If the network architecture is fixed, the bandwidth usage can be modeled and algorithms can discover strategies to route information in the fastest and most optimal manner across the network. However, the space-turtles will be moving in deep-space, subject to the relative gravitational pull of different planets and stars. By the time a packet is received by a space-turtle, it may be at a completely different location in space than when the packet was transmitted. This requires the development of a new generation of network routing algorithms.

We also need a new generation of error-checking and parity algorithms to ensure the correctness of the transmitted message. This is even more important for software updates in deep space, where the correctness needs to be checked also in the temporal context i.e, is this the most accurate software for the current physical location that the space-turtle is located in ?

But the beautiful thing is that all of this algorithmic innovation can happen in parallel on earth, independent of how and when the space-turtles are deployed. We can build such algorithms and test them in simulation, or through practical applications such as robots exploring the deep oceans.

Ultimately, a modular and organic network of space-turtles will  decide where to flock together and exchange key pieces of information with each other. This can happen in locations where the space-turtles can recharge their batteries, for example, using energy of the stars. Later on, as the space-turtles explore inter-galactic spaces, they need to communicate with each other to meet at a specific location in thousands of years of time.

I think it is informative to look at the flocking behavior of turtles in the ocean to understand how space-turtles should meet with each other. Turtles get together in large numbers on some safe islands to spawn the young. These young turtles suffer an extremely high mortality rate, but some of them survive and continue the journey. Ultimately, these explorers will have to work in tandem and  identify a planet that is the most suitable for recreating earth-like life. This will be a journey that takes millions or even hundreds of millions of years. But that is okay, we have ample time until the heat death of this universe.

Encoding DNA for millions of years: 

DNA can now be sequenced and synthesized

So far, we have not talked about how to transmit life. There is no life-form on earth that can survive such long journeys in a single life span. Some of the hardiest species - the extremophile Tadrigrades - may survive for decades in a low energy hibernated state, but not any longer. The living tissue of organisms will also need to be protected from high energy bombardment by cosmic radiation. In this scenario, it is foolish to even ask if life can be transmitted in vivo to another galaxy.

However, there is a different possibility, involving the recent technologies of genetic engineering and synthetic biology. It is now possible to write down synthetic DNA into a cell and store it for hundreds of thousands of years. With improved storage, this DNA can last for even millions of years. In fact, this is one of the most promising methods for archiving media information. At an annual company conference last year, when I was working at Technicolor Research, I had the singular privilege of looking at a movie stored as DNA in a colorless solution in a tiny test tube. Using the latest gene sequencing technology, in collaboration with Harvard University,  researchers converted the pixel values in video to a code defined by the nucleotide pairs in DNA. Information is information, and can be encoded into any language and decoded back. In this context, the language is made by the two base pairs of DNA: guanine-cytosine and adenine-thymine. We can create a code to represent the video film and transcribe it into synthetic DNA and thus archiving it.  This is by far the best archival format that is currently available for movies. If we can archive a movie, we can also archive the entire genome of Homo Sapiens. Indeed, we can archive the genomes of millions of species that are necessary for our own survival.

At the other end, there is a decoder, which is a gene sequencing machine (these are already becoming extremely cheap). So far, it is good enough for decoding the video of a film and to project it onto a screen.  But if we want to create life, then things are more complicated. At the current state of technology, we need a frozen egg cell into which the DNA can be impregnated. Our current technology cannot freeze egg cells for more than a few decades.  But this might change in the future.  Another possibility is to synthesize egg cells from stem cells. This is also a technology undergoing rapid development. With the right type of storage and synthetic biology, the decoded DNA can be impregnated into a living cell and then reconstructed into an organism. If and when such technology develops on earth, it can also be deployed on a distant galaxy, a million years deep into the future.

When we have this technology, we can deploy it within the space-turtle. Then each space-turtle contains valuable cargo: gene sequencing machines, DNA storage of genomes, frozen stem cells that can transform into egg cells, as well as computers and navigational instruments. Another tantalizing possibility is that the space-turtles mimic the process of reproduction itself in space. Like how a mother turtle gives birth to a baby turtle, a space turtle should be able to 3D print a baby spaceship in its exact make. During the period of gestation / pregnancy, the mother space-ship can expand in size, making room for printing out the parts of the baby ship. The raw material for 3D printing can be provided beforehand, making it sufficient for producing several generations of space turtles. Another possibility is to collect raw material by grabbing cosmic dust. Over hundreds of thousands of years, sufficient cosmic dust can be accumulated to produce a child space-ship.

Thus, these space-turtles will build a civilization of patient and peaceful life-forms, entirely robotic in origin, but harboring the hopes and DNA of biological life on earth. They will roam distant galaxies in deep space for millions of years. Together, they will build a detailed map of the universe with its myriad galaxies and mysteries. They may even do some scientific investigation on their own, and continue where we humans left off. But ultimately, they will find a suitable habitat for life - the kind based on DNA and wet biology  - so that we can together ponder even greater mysteries.


Reconstructing a biome:  

Species interconnections in a biome

If we take millions of years until we find a suitable planet, we may also take millions of years to make it inhabitable for ourselves. In fact, we can run a fast-paced simulation of biological evolution on this planet. We can accelerate specific evolutionary turning points, by turbo-charging the DNA of the species. As life proliferates by harnessing energy of the stars, it will also make significant changes to the atmosphere of the planet. For example, cyanobacteria can create Oxygen on the planet's atmosphere. Huge forests and coral reefs can be created. These will support large biomes of interconnected species. Predators and prey may flourish for millions of years on this planet and undergo some biological evolution themselves. The space-turtles may keep visiting the planet at periodic intervals, to check the progress on how the planet is yielding to life.

As and when the conditions become appropriate, the space-turtles will introduce key species onto the planet: those which we consider as the fondest memories from our planet earth.  Ultimately, they will introduce humans as well. This may take several millions of years.

I think the current state of scientific knowledge is more or less sufficient for the problem of space exploration, but definitely not enough for reconstructing a biome by modeling the ecological interconnections of species. But hopefully, our knowledge will get better in the coming centuries, and the space-turtles can update themselves over the intergalactic network.

In addition to ecological knowledge, we will also upgrade our psychological knowledge on how and when  environment affects the growth of a person. If we are able to identify the different life-changing moments in a person's life and capture them in the digital format, they can be encoded along with a person's DNA. These phenomena in the environment can then be reproduced exactly in that order, which will create a replica of the same person that was living on earth, several millions of years ago. Thus, we will achieve a version of immortality, as well as a possibility to learn from the mistakes in a past life. We can avoid the terrible mistakes we did as a  civilization, but also those that we did as individual human beings.

If this is not a plan for achieving paradise, I don't know what is.

Relay stations of rocketry: 

In this blog post, I talked about a quasi-spiritual dream. I read in the book "Adventures in Human Being" by Gavin Francis, that humans who keep smiling a lot will have a physical imprint of their smiles on their faces as they grow older. The facial muscles responsible for the smile, when exercised a lot, leave their imprint not only on the physical appearance of the person, but also on the brain. The nerves connected to these muscles produce happiness hormones and make the person think bright and happy thoughts.


So if we keep trying to smile, we may indeed become happy. I think the space-turtles are such a dream that induces a smile. It is far-fetched, but it is not impossible. As we try to build them, we may indeed change our society. By forcing ourselves to think long-term over millions of years, we will see the pettiness of today's politics. We will feel the outrage caused by man's destruction of the nature's treasures. We will be ashamed of the  sacrifice of millions of human lives for nothing. We may even put an end to the endless wars of greed. In this regard, space-turtles are a story I want to believe in. It may not be true, but it is a story that will lead us to a better world.

I have left one thing left unsaid: the whole saga of the space-turtles starts by building a factory outside the earth's atmosphere. This narrative gimmick let me short-circuit the entire process of sending a space-craft up in space. At the moment, only large nations (and billionaires like Elon Musk) can send a space-craft to outside the earth's atmosphere. What is the likelihood that anybody will support my plan ? Zero.

So I need a reasonable plan for this first step: for overcoming gravity and reaching beyond the earth's atmosphere. The computer scientist Alan Kay (one of my heroes) has once said that  "moonshot" goals are terrible for the progress of science.  For example, the Apollo missions have sent humans to the moon, but all space exploration collapsed after that. It became simply too expensive to pursue the leads obtained by the technology. Instead of moonshots, what we ought to have developed was a scalable and organic model of space exploration. I think the space-turtles yield such a model, that grows naturally and ultimately scales to an intergalactic network. But can we build such a model also for rocketry and to reaching beyond the earth's atmosphere ?

If we break the task of overcoming earth's gravity into small chunks, then individual humans and companies can solve them. After all, many companies are capable of making airplanes and many individual humans are capable of making robotic drones. These flying machines can reach great heights, but they cannot refuel in the air and make pit-stops until they  reach beyond the earth's atmosphere. But why not ?

One option is to build a space elevator - a long cable tethered to the ground through which energy can be transmitted. But this again requires huge investment from nations, which may not be forthcoming. Another option is to build flying pit-stops, where other planes can charge and refuel.

If we tether some equipment to a large Helium balloon, it can already rise beyond the earth's atmosphere. But there, it deflates and falls down to the ground. May be, it is possible to coalesce material together in the higher reaches of atmosphere. Like pieces in a jig-saw puzzle, teams of drones can dock into each other and build a greater space-ship. Such navigation and docking is not impossible. With today's technology, we can already perform highly accurate navigation on roads and stormy weather. Similarly, such intelligent navigation technology may enable multiple drones to find and dock together, after which they spend a small amount of fuel and reach into a low-earth orbit.

Another possibility is to transmit energy by laser beam. Each of the drones can stock up some energy, and beam it via Laser to a mother-ship, which can use all of this energy to overcome the earth's gravitational pull and reach a low-earth orbit. There may be failures, some of the drones may fire the laser beam inaccurately. But the nice thing is this system is scalable. It will be possible to repeat the experiment and learn from the previous failures.

Ultimately, if and when life escapes from this planet, it will do so using methods that are scalable and organic.  Like the space-turtles, we need to think in terms of millions of years. Achieving a relay system for rocketry is probably not that hard. We may even solve it in just a few hundred years.

Keep smiling, space-turtles !