Internet in Space

Have you ever asked yourself "Can I stream cat videos if I move to Mars"?

Well, maybe not, but with initiatives like Mars One, and visionaries like Elon Musk, who claims that he's going to put people on planet Mars in 2026, the question about how the internet is going to look like as it expands from planet earth and into space is going to be important. Is it going to be a shared infrastructure, or is it going to be owned by a company or country? Because as we take the first steps towards inter-planetary communication, a shared and decentralised approach similar to how the internet is setup today would be both desirable, and feasible - as I will describe below. In fact, the structure of this technology will be critical, because it will define how we will collaborate, and thus, how far into space we will be able to reach, in a similar way as language defines how far we can think. 

Artistic rendering of Mars One colony

Image: NASA/JPL

InterPlanNet

Vint Cerf: The Father of the internet 

As an example on how important the structure of a technology is, consider how the internet was designed. Much of the success of it has been thanks to how it was specified by - among others - Vint Cerf - in the 70's. The explosion of collaboration and creativity that it has enabled is much due to the fact that it was designed to be inclusive, open, and decentralised. So it is great to see that Vint Cerf is working on the next inter-planetary version of it, called the InterPlaNet.

What is significant for the current era of space exploration is the privatization of space, and the increasing cooperation between private space companies and governmental agencies. One example is the news last week about Boeing and SpaceX getting NASA contracts to take astronauts to the ISS. This commercialisation of space in combination with commercial trends like crowd funding and resource sharing will drive the need for more standardisation and open protocols in space.

Me at the small-sat conference next to a model
of SpaceX's "Falcon Heavy",
which will have a payload of up to 50tons.

Below are a few topics connected to the future evolution of space-based communication:

1. Crowd-funding of satellites and data:

The dropping costs for launching satellites combined with the power of crowd-funding will give rise to a new constellation of nano-satellites which will provide broadband access to people and organisations in remote areas. An example could be that Red Cross sets up a kickstarter campaign where people can donate money to support Red Cross in Kongo with X GB of data to help remote diagnosis of patient in one of their remote hospitals.

21st of September 2014 : Launch of SpaceX's Falcon 9. 
Launching satellites as "rideshare" on secondary payload drives down costs for nano-sats

2. Interplanetary Communications and Latencies:

As we start to explore off planet colonies we will need a reliable inter-planetary communication network. One parameter that will shape this network is the long latencies. The minimum latency between Earth to Moon is 1 second, but between Earth and Mars it ranges between 3 and 22 minutes. This will make re-sending of messages painful, so the link between planets would need to be asynchronous, and very robust, with path- or data redundancy, and built in checksums. The latencies will be a big problem, and make it very difficult for desirable telepresence applications, like remote controlling robots and spacecrafts on Mars, so we will need to completely re-imagine this kind of communication, possibly using AI for enabling "simulation modes" which cuts down data, and gives the impression of shorter latencies. 

3. Dynamic Routing and the need of resource sharing

As for the physical network, the fact that the direct line of sight is often broken - due to planet's rotation on its own axis - will make it necessary with adaptive routing of messages (similar to how the internet works). Due to the scarcity of nodes in deep space, resource sharing (describe later) will also be important. An illustrative example is how the Mars Rover was routing it's messages via the two existing Mars-orbiters to the DSN-antennas on Earth of which one out of 3 would always be in line of sight with the orbiters. Such a setup indicates the need of re-using resources in space. A dynamic mesh-network of spacecrafts from NASA, ESA and Roskosmos as well as commercial players could help routing more efficiently between planets. With recent initiatives around astroid mining like Planetary Resources, and NASA's OSIRIS-REx , astroids might also be used as nodes of such a mesh network to help supporting uninterrupted communication between different celestial bodies even through out eclipses. 

Increased resource sharing could also be beneficial for Earth-based satellite communication. For example between GEO- and LEO satellites, where a routing network of GEOs can provide a backbone where nano-satellites in LEO can provide better reception from low-powered mobile terminals.

4. Resource-sharing of space equipment:

Similar to how AirBnB enables people to share under-utilised resources, an intelligent meta-network could help maximise utilisation between resources in space, such as communication satellites. This would help commercial interests to maximise return on investment of space equipment. Sharing of resources would make sense from both consumer perspective, where users don't want to tie in to a specific network / constellation due to fluctuation of usage and uncertainty of coverage, and also from the perspective of resource owners, where your satellite capacity is most likely under-utilised in remote areas.
This need for resource sharing will help push open standards, and open networks, and also incentivise the standardisation of protocols as well as space hardware. This is already happening with SPA (Space Plug and Play Avionics) which is a cooperation between Swedish ÅAC Microtec and US Airforce.
¨

The potential future of SCaN: optical links, interplanetary Internet, and real-time video feeds.
Image source: NASA

If the infrastructure is constructed in the right way, the benefits of sharing infrastructure will be obvious even for mission critical communication elements. Here are a few ideas around things that could accelerate the adoption of this:

Physical Resource Monitoring: An open system for monitoring satellites positions and orbits.
Address Protocol: The development of a standardised protocol for describing, and addressing different resources in space.
Quality of Service: Since the owner of the physical link will want to ensure guaranteed bandwidth for mission critical elements before allowing sharing of infrastructure. Thus a form of QoS will be needed. This QoS would also need to take into account specific constrained resources like power and heating.
Security: need to be good enough for mission critical elements, while still be possible to decrypt onboard a spacecraft.
Decentralization: Just like the original setup of the Internet, this network needs to avoid the vulnerabilities that comes with one centralised control point. The adress-space as well as identity, and authentication schemes would need to be completely decentralised. Maybe inspiration could be drawn from the block chains used by BitCoin?

5. Onboard Computation

As computing power increases exponentially we see an increase of onboard processing of data. One example is the hyperspectral remote sensing satellite that we have been investigating at Singularity University with our project nuTerra, where we consider using a radiation hardened AMD-chipset from BAP which fits in a 1U CubeSat, and gives you the same computing power as in a Playstation 4 with 2 watts power. 

Marcus, Brian and Angela from nuTerra
with BAP's radiation hardened AMD-board 

This will help doing analysis onboard, and deliver ready-analysed results in realtime without having to download the full data to ground-stations for analysis, and thus cut the delays associated with such transfers. With a good sharing infrastructure in place one could also consider the evolution of specialised "cloud computing satellites", which are optimised for handling fast computing to other satellites. The use of Free Space Optical (FSO) communication between satellites will enable increased inter-satellite bandwidth. Computing like this might also be centralised on ISS, or distributed to GEO satellites, where power and temperature is less of a problem than onboard nano-satellites.

The Voyager Golden Record

6. Communication channel for extra terrestrials:

In the event of a contact with intelligent lifeforms outside of this planet we would need a globally accessible relay for enabling open communication. It would need to be carefully designed to our best effort to allow both security as well as transparency, and be truly globally accessible. We can't just ask the ET's to go and sign up for a gmail address! ;-)

Let's make sure that our networks stays open, and inclusive, because the structure of this technology might be just as important to space travel, as DNA has been to our evolution.


And just think about it! Internet+InterPlaNet+Lasers+Space+Cats=The future 

Clean Disruption of Energy

Today's lecture by Tony Seba - author of the book "Clean Disruption of Energy and Transportation" painted a fact-based and optimistic view of the future disruption within energy. In short, Tony argues that by 2030 fossile fuels will be abandoned in the same way as the salt-mines of Salzburg has been abandoned due to new ways of producing salt. I had a short chat with Tony after the lecture, and he explained the theory more in detail. Essentially there are 2 factors that drives this, and that is:

1. The price of solar panels are dropping every year, and efficiency is increasing, which eventually will make solar electricity very cheap. One of the biggest impact to the adoption of solar panels has been innovation in business models, for example Solar PPA's done by companies like Solar City, where private persons can sign an agreement to lend their house-roofs to Solar City which will provide, setup and maintain solar panels. This means that private persons does not need to take any risks or have to front the initial investments of installations, and can start saving money from day one!
Currently - according to this article from IEEE -  the price/performance of solar panels are on pair with the price of electricity from the grid (Grid Parity) as far north as in Holland.
IKEA is powering 89% of their shops with solar panels in the US (source)!
Some claims that productions of solar panels consume a lot of power, but Tony sais that First Solar claims that their panels recover their energy payback in 9 months. The same time for power payback of a nuclear powerplant can be as long as 14-18 years according to some studies.
Tony also mean that in 2020 - for the US Southwest - the cost of solar power will equal that of transmission of power through the power grid (cables, poles, etc). This means that even if grid-power would be FREE, it would be cheaper to go solar. Details can be found in this video

2. The price of power storage has fell by 16% since 2010, and this trend points to that Electrical Vehicles (EV) will drop in price to a point where it's just not feasible to run on gas anymore. The recent news of Tesla open sourcing all of it's patent is also driving innovation in this field, and will help lowering the costs of EV's.

Tony also talked about nuclear power, and said that it's clearly so that the spread of nuclear weapon material in the world is due to the spread of nuclear energy. There is no clear boundary between nuclear power and nuclear military industry, and some countries which invests in nuclear power actually wants a viable 'civilian' nuclear ecosystem to maintain their military nuclear options. I.e. producing nuclear power for the ability to have nuclear materials and rest-products for bombs, and the infra structure and know-how on how to assemble one. Examples of this could be Japan, US and France.
The cost of nuclear power has increased 10x since 1970, so nuclear in itself is not economically justified. There are also an interesting study published by Versicherungsforen Leipzig in Germany one month after Fukushima, which claims that IF a nuclear failure where to happen in Germany, this could cost 5.5 Trillion Euros! And there is NO insurances that covers this, which means that the government (and in the end the tax-payers) are paying the insurance based on the Price-Anderson act

So, the energy question is clearly political, and Tony points to an example of this from India: According to the IMF-report the subsidiaries of fossil-fuel is mostly benefiting the top-earners, even though the  arguments for the subsidiaries is that it should benefit the poor. In the IMF-report it has been proven that these subsidiaries are mostly benefiting the industries, and in the end the very top of the population. If India would spend the same money building residential solar panels, this would provide electricity to all those 500 million users that currently does not have any electricity at all. At no additional costs for electricity production.

And why is this important?
Because 7 million people die every year from the effects of fossil fuels (mostly due to air-pollution)
Aftermath of Fukushima has caused cancer in around 200'000 children

More on solar power:
Grameen Shakti and their micro credit institute has managed to help 1.3million homes in rural areas to be run on solar panels

To be continued...

First day at Singularity University

After 5 years silence, I've re-purposed my old "general purpose" travel blog to use for writing about what I experience here at SU.
It's sunday and we've had a first round of introduction,  and it's amazing how many intelligent people is gathered here at NASA's research facility in Silicon Valley. We also did Peter Skillman's "Marshmallo challange", and managed to build 13.5inches tower.

(Winner managed 29 inches).
Paulo (om my right) is my room mate here at SU, and we're both aspiring DJ's!!

But, here are a few random ideas that we've been discussing so far:
Space elevator & astroid mining:
A gigantic space elevator could make astroid-mining possible, and trigger the world of abundance, and inflation similar to what happened when the Spanish brought gold from the "New World" to the "Old World" as described in the book "The ascent of Money".
The lab...
here at SU looks very promising, and i've only scratched the surface by playing around with the Oculus Rift, dissecting a spider with Leap Motion controller and discussed the problems with 3d-interfaces and "confirm actions" (clicking) with the lab assistants.
Body tracking:
I learned about the Scanadu which is a medical device (1-min overview here), and we've been talking about how new non-invasive sensors for monitoring your body could be combined with personalized nutrition supplement to help you perform better.