Before answering this question, let’s take a few moments to look at what a robot is. Whenever we think about a robot we usually imagine a human-like machine with jerky movements and stilted speech. Something like Robin Williams in “The Bicentennial Man”.

This is the image we have been fed by science-fiction films and novels. In fact a robot is a system that functions autonomously or by remote control, especially any machine that can be programmed to carry out tasks normally performed by humans. This definition includes a great number of devices, ranging from electronic food processors and robotic vacuum cleaners to simulators, space rockets or space robots (WALL-E).

But can a robot make decisions in its own right? Or can it only follow pre-programmed human instructions? Although all of us who have seen Terminator might expect the answer to be “no”, the fact is that work on decision-making robots is now making great headway. This doesn’t mean they can deviate from the “established plan” but the orders can now be given to them in a vaguer way, more similar to how people speak to each other. For example, it is much more useful to be able to say “come here” to a robot ten metres away in an obstacle-strewn environment, the robot then finding its own way towards us, than having to guide it along each step of the way: ‘look straight ahead, take a photo with the lefthand camera, take a photo with the righthand camera, get a stereoscopic image and check if there are any obstacles, lift up your head, now lift the right leg, lift the left leg, etc.’

In GMV we are delighted to have participated in the development of a software controller that can be used as the “brain” of one of these more autonomous robots. As a GMV engineer I’ve formed part of the team working on the GOAC project (Goal-Oriented Autonomous Controller). This is a two-year European Space Agency (ESA) project that aims to develop an intelligent controller for space missions with the participation of robots capable of making some decisions in their own right.

The use of robotic systems for space exploration has opened up possibilities for solving certain problems that have plagued space missions for some time. On most space missions there are long round-trip communication delays; this means that depending on the ground for decision making might be detrimental to mission performance, especially in terms of scientific operation and optimization. The spacecraft’s safety might also be jeopardized in environmental conditions of high uncertainty, as usually occurs when exploring the surface or atmosphere of a planet.

This situation can be improved if the space segment is based on a robotic system capable of deciding for itself. The robotic system can perform mission objectives in a much quicker and safer way by closing the decision-making loop on board rather than checking back to ground each time. It can also cope with scientific objectives that crop up during operations. Freed of these petty but time-consuming tasks, ground operators and scientists can focus more single-mindedly on their own work. This can happen only if the robotic system is highly autonomous, i.e., capable of fulfilling its objectives without outside help.

In this particular context autonomy means a reasoning ability: the operators tell the system what to do without getting bogged down in details and the robotic system itself will then decide how to go about it.
To be able to make decisions in its own right a robotic system had to be fed with information on all possible behavior, including all types of operational and safety constraints, and it must be capable of continually perceiving environmental conditions.

The reasoning ability has to be accompanied with a safety mechanism preventing the robot from carrying out certain actions. In GOAC we have used a technology that “imprints” a series of unbreakable rules on robot behavior, i.e., forcing it to behave correctly when carrying out a certain task. This mechanism acts in real time without needing to trigger a reasoning mechanism, which normally requires a certain lapse of time. It plays a similar role to our reflexes: these act “instantly”, bypassing the higher-order capacities of our brain.

Although system development needs to be completely foreseeable in some scenarios, increasingly autonomous robots are likely to be feature of the near future, especially in space exploration.

Please feel free to give your opinion below.

Author: Antonio Ceballos

Last month, on 20 January, two of our colleagues were interviewed by the National Spanish Radio about space debris. Following on from this event we are going to publish a series of blog posts on this burning issue.

MEDIA INTEREST

In recent months uncontrolled satellite re-entries have hit the news three times. The first was the American satellite UARS which re-entered the atmosphere over the Pacific Ocean on 24 September 2011. Shortly afterwards, on 23 October 2011, the German X-ray telescope ROSAT did likewise over the Bay of Bengal (Indian Ocean). More recently, on 15 January 2012, the Russian satellite Phobos-Grunt, originally programmed for a complete Mars mission, careened down over the Pacific Ocean.

On their re-entry through the atmosphere these bus-sized objects break up progressively into smaller and smaller fragments, some of which might reach the earth’s surface. Nonetheless the likelihood of one of these rogue objects killing anyone is remote; we should not forget that most of the Earth is covered by water and there are also large land areas that are uninhabited.

As well as these large satellites, other space junk is continually re-entering the atmosphere, though it normally breaks up completely on the way down without posing any threat to people.

At present there are about 1,000 operational satellites of various types in orbit (telecommunications, navigation, earth observation, scientific and military research). As well as these missions there are another 15,000 to 20,000 listed and monitored uncontrolled objects in space. There are also about 300,000 objects ranging from 1 cm to 10 cm in size that could destroy another satellite, and millions of even smaller objects (less than 1 cm) that slip below the radar but still pose a huge risk. All these earth-orbiting objects not belonging to active missions are lumped together under the term “space debris”. Space debris is made up by objects as varied as large rocket chunks and old satellites, remains of explosions or collisions and tiny particles.

PROBABILITY OF PHYSICAL HARM

As we have already pointed out, many objects re-enter the earth’s atmosphere every day. Most break up on re-entry and never reach the ground. Few are the objects big enough to survive the whole break-up process and actually come to ground.

Even debris that does not break up completely on re-entry is very unlikely to cause physical harm to anyone on earth (odds of 1 in a million) and there has only been one known case of impact with humans to date.

PROBABILITY OF COLLISION WITH A SATELLITE

The real risk, however, stems from the probability of collisions between different objects in space. The upshot of such collisions might be mission loss, exponential growth of space debris and consequent increase of the collision risk in a vicious circle, which could in last instance make it difficult to access the Space environment in the future.

The likelihood of collisions between two space objects is still low but is continually increasing with the growth of space debris itself, especially in the more saturated areas (low almost polar orbits, from 800–1000 km high, and equatorial and geostationary orbits at a height of 36,000 km).

Space debris is a growing concern. The more uncontrolled objects that are out there, the greater the likelihood of catastrophic collisions, causing not only the loss of one-off missions but also a collisional cascading chain known as the Kessler Syndrome. This would exponentially increase collision risk and pose a serious hazard for future space missions, especially manned ones.

Satellite operators are therefore becoming increasingly alarmed, especially after the collision between an operational satellite of the Iridium constellation (satellite telephony) with a decommissioned Russian satellite in February 2009.

Consider these telling facts: NASA now has to make over 10 satellite maneuvers a year to avoid collisions; ESA has to maneuver its most important satellite ENVISAT more than once a year and the International Space Station has had to be maneuvered and the crew evacuated to a safety pod on several occasions due to space-debris near misses.

DETECTION, OBSERVATION AND MONITORING OF SPACE DEBRIS

The state of space debris is monitored from space surveillance systems that detect and keep an eye on orbiting objects. This involves powerful radars (for near earth objects) and telescopes (for higher orbits). These systems can list and monitor orbiting objects above a certain size (over 10 cm in low orbits and 1m in high orbits).

Several countries now systematically monitor space debris:

• The USA runs a worldwide military Space Surveillance Network to list these objects. Part of this information is published (sometimes of a degraded quality and never for missions considered to be critical to its interests).
  
  
• Russia also has a (more limited) space surveillance capacity but publishes no information on the matter.
  
  
• Since 2009 Europe has been working busily towards its own Space Surveillance and Tracking capability. Spain as a country is currently the main contributor and its space industry is playing a crucial role in developing the system. GMV in particular is participating in and leading several ESA projects in this area.

In upcoming posts I’ll be talking about how the space debris problem might be mitigated and the junk removed.

I hope this has been of interest to you. Feel free to leave your comments below.

Author: Alberto Águeda Maté

Short battery life is one of the commonest complaints among users of portable computing devices. In most cases a top-range laptop offers from 5 to 7 hours of real use. The latest rumours about a solution to this problem are based on a patent filed by Apple for fuel-cell-powered MacBooks with a much longer battery life.

Unlike a standard laptop battery, fuel cells are electrochemical devices capable of generating electricity from the reaction between an external fuel source, normally hydrogen, and oxygen. Fuel-cell powered laptops, if this idea comes off, could run for days or even weeks between charges. Overcoming the limitations of traditional batteries would also make them even slimmer and lighter.

Apple has been trying to improve its products energy efficiency for quite some time, pursuing an active green IT policy, although we are unlikely to see commercial fuel-cell powered products in the short term. Several of the sector companies like Toshiba, Samsung and Sony have been working for years on the development of these products and have showcased tradefair prototypes though none has yet made the leap onto the shop shelf.

Author: Jesús Mariano Pascual Díaz

One of the recurrent problems of using cloud computing services is law abidance, since, due to the very architecture of cloud computing, data can slip back and forwards so easily between countries with different laws.

This was particularly the case for personal data, which is protected in Europe by specific legalisation that limits the exporting of this data. Problems were also posed in the US with the cross-border access granted under the Patriot Act (http://www.backup-technology.com/9796/u-s-patriot-act-dampens-microsoft-cloud-services/). Some providers therefore opted to set up several cloud services distributed in different geographical regions, adapting each cloud to the local laws in each region and setting up a specific cloud for Europe (e.g. Amazon); in other cases there have been government initiatives, as proposed in France, to set up national cloud services adapted to its own laws (http://www.elpais.com/articulo/Pantallas/nubes/Internet/crean/ problema/soberania/elpepugen/20110927elpepirtv_2/Tes)

Microsoft has taken a different approach for its Office 365 service. It has decided to apply European data protection laws to all its clouds supporting the office365 service. To do so Microsoft decided to apply the Model Clauses derived by the European Commission from the European Data Protection Directive; these clauses lay down personal data protection measures. It has applied these clauses in all its datacenters running office365. To put it another way, Microsoft has voluntarily decided to apply European Law outside Europe. Thus, the data of a Spanish user on office365 might be run in centers outside Europe (USA; Japan, .... or wherever) while complying with directly applicable Spanish law.

It has also set up the Office 365 Trust Center trust.office365.com, which gives information on its information processing principles and additional details on the employee visibility of client information.

To round out the story, all we need now is an official communiqué from the European Commission on this matter. After all, it’s one thing to claim you are Beyoncé’s friend and quite another for Beyoncé herself to say "This is my friend."

Author: Mariano J. Benito

From GMV we want to wish you happy holidays and a happy new year.

This drawing is the winner of GMV's internal greeting card competition 2011. The author is Nerea Arbildi, 5 years old.