Thinking Olive Tree

Personaly I think you are wrong

2011-07-08T14:28:00.000+01:00

I saw this video here, on the IEEE Spectrum blog. The bottom line is that soccer robots are evolving fast and soon will beat the crap out of human teams. Well, I don't doubt that will happen sooner or latter. After all, a lot of effort and money is going into making robots play soccer.

However, notice that in the fatal day that robots beat humans in soccer all that it proves is that you can built machines to beat humans at a particular task. No shit Sherlock ! We see that in our everyday lives. Last year we saw a computer winning Jeopardy and a couple of years ago a computer won a game of chess. There is nothing extraordinary about this, although they are very impressive technical achievements.

Personally I am convinced that it is impossible for machines to operate autonomously as living creatures having been doing for ages on this planet. And this places an upper bound on what machines, robots for instance, can accomplish no matter the amount of geeks involved in building them. I look forward to having robots beat me in soccer (not that hard given my growing beer belly) but I honestly doubt "we humans are doomed" as is implied by the author of the blog post.

Time is scarce (still)

2011-06-20T17:53:00.000+01:00

I got to read this post in the blog Everything Robotic, which raises the question of why don't the manufacturers of industrial robotics move into the area of service robotics.

It is argued that the business model that these companies are accustomed to is not suitable for area of service robotics. I agree and would add that the way innovation occurs in both areas is also different. For industrial applications, the ideas have to be (i) patentable and (ii) money cash cows. The two are related and that very much guides how business is conducted. With proprietary solutions, a lot of special training sessions for workers and you get the client by the balls. The manufacturers have to guarantee mean-time-between failures and upper bounded repair times. However I am sure that somewhere in the fine print of the maintenance contracts is written "Hey, shit happens. Deal with it".

In any case service robotics is much more about the application, not so much the tools to get the job done. That is why I believe that is it much more difficult to sell mobile robot solutions in this area. Basically, we are getting along just nice without mobile robots in the office, the house or the garden so why should we buy a fancy machine ? Especially now that people are cashless and have a crisis knowing at the doorsteps. At least those of us in Southern Europe.

I am not as confident about garage start-ups that sprout from research centers for the simple reason that the academic world is much more about nice theoretical concepts than actual working robots. But I might be wrong.

A new distro of Ubuntu comes out tomorrow

2011-04-27T22:48:00.000+01:00

Bought a new laptop today. Nothing fancy, just a budget laptop for my work. By default it comes infected with Windows, the seventh incarnation. It took more than one hour to run the automatic setup of the operating system and the bloatware it ships with. Man, one hour to setup. Did I mention the new Ubuntu distro is out tomorrow ?

Time is scarce now

2011-02-17T15:59:00.000Z

but a few things have caught my attention in this past week:

Cubelets are just cubes which you can snap together and immediately begin interacting with each other. Cool idea in bright, flashing colors. Think of them as lego bricks on steroids. They are sold as a "modular robotics kit", which can be great fun for kids and grown ups alike. But that is just that. Not sure how many different types of cubelets there are, but I have the feeling that there aren't all that many different combinations you can try. I would not put that lego Mindstorm kit away just yet.

The RobotEarth is a project aiming at collecting the learning experiences of different robots and making them available to every cybernetic creature. Think Wikipedia for robots. The motivation is to try to avoid re-inventing the wheel whenever a robot must solve a learning problem. And that is just fine, but just think about the countless different shapes robots come in. Sensors, actuators, hardware, colors and smells. Ok, that last one is streching it a bit. And now think on how alike humans are. No, really. Despite external physical appearances, and some internal ones as well, humans are very much alike wherever you find them. Unlike robots, which pretty much came in all shapes and sizes. That is why I am a bit skeptical on the utility of this robot aimed Wikipedia. But hey, after so many encounters with the robots in the field one cannot avoid becoming a cynical pessimist.

Finally, the Bilibot. If it ever manages to sell, it promises to be robotics on a shoestring. Well, at about 700 US$ it is still an expensive shoestring but at least they aim at making it available to the masses. Let's hope mass marketing can make it cheaper.

Summary of last week

2011-02-08T09:25:00.000Z

the "another one" being me.

Post-battle state of mind

2011-01-31T19:20:00.000Z

Spent all day battling the evil hardware gnomes

2011-01-31T16:47:00.000Z

... yeah, I lost.

Evil hardware gnomes 2 - Me 0

2011-01-26T14:18:00.000Z

For the past three weeks I have been trying to get the robot to navigate autonomously in the university campus. To be more correct, I have been trying to get a worthless, overpriced piece of ill-designed junk built by MobileRobots with the sole purpose of tormenting PhD students, to navigate in the university campus. I spent one hour connecting to the onboard 5 V, supposedly easy to access. However, the whole process makes open heart surgery look like kids play !

Mail ? That's so 1980's...

2011-01-16T23:07:00.000Z

The magazine Business Week has an article about office robots becoming common place in the nearby future. Among other things these robots are going to deliver mail, yes mail, you read it correctly and pour coffee for their "co-workers". To be more precise, they mean this bad boy

Kawada HRP-4

quite cool, right ? Too bad the price tag is about US$ 350,000. I do not know the salary of an office page in Japan, but if it is about US$ 30,000/month then I am packing to Japan. Now !

Another company in the USA, SmartRobots, is selling a much cheaper toy, sorry office worker. The SR4 it does not come cheap either as it sells for a little less than US$ 19,000. It is not nearly as cool, as the japanese bipede toy, but beggars can't be choosers. And it gets points for the pure geek aspect of the thing.

SmartRobots SR4

I know that journalists are not famous for their brain power, but doesn't the mention of a mail delivering robot even manages to raise some eyebrows ? Come on, mail !? And on top of that, the other task is puring coffee !? Really ? Why don't you just lock the employees in their cubicles while you are at it ? Sometimes I feel that I am the only intelligent person left on planet Earth...

Some facts are hard to digest

2010-12-06T17:00:00.001Z

thank you xkcd.com

We are asking the wrong questions

2010-11-26T00:38:00.001Z

so it is only natural that we get the wrong answers. At least this is what I believe to the main reason behind the field of mobile robotics successively failing the wonderful predictions that were made. That and the not so unreasonable hypothesis that maybe robotics is close to the border of what it can really do. To put it in another way, what we have right now in terms of robotics technology and achievements is pretty much what we will have in the future. Sorry, no 2CPO or Captain Data but maybe an R2D2 is possible to come by in the future.

And what are the questions we are asking right now ? Take for instance, the standard navigation task: robot must move from A to B. Technically it is more complex than this, but you get the point. The general approach to solve this problem is to ask for the optimal path, in terms of travel time or distance, that gets the robot from A to B. The problem is not as trivial as it might seem at first, and it can be pretty tough to solve. But it can be solved, provided you have enough money and time to spend. As with any engineering problem.

Now, let's pause for an instance. The problem is being formulated in a very sound manner. Get me the optimal path that takes my cybernetic creature from the bedroom to the living room. Along the way, don't forget pick me up a beer from the kitchen fridge.

Is this the way you go at the navigation task problem, when you need to solve it ? The question is not easily answered because you are not aware of the unconscious thinking processes that take place in your brain. So let's look at some of the most computationally handicapped creatures God placed on this Earth: insects and birds. Dolphins and mice don't count because, as you know, they outrank us intelligence wise.

Take the route traveled by an ant or a small bird, while foraging for food. Is it plausible to assume that these carbon-based life forms compute an optimal path, while moving in the environment ? Certainly not, since that would require at least a reasonably accurate representation of the environment and the means to reason within that representation. In particular, birds move in 3D space which is a lot trickier than the almost 2D spaces humans move in. Not to brag, but if even humans have difficulty in getting anywhere on time how can an ant or an hummingbird plan the path it will follow ?

Of course, these creatures might used simple and efficient algorithms, finely tunned over thousands of years by evolution, that implicitly take our crawling friends along the good road. However, using an algorithm that might produce a near-optimal path is not the same as actually computing it beforehand or while on the move. So instead of asking for the optimal path, which a nice mathematical way of showing of just how geek can roboticists be, we should be asking what are the efficient algorithms that generate good enough roads.

Now you might ask, if that is so how come nobody has thought about that before ? Well, they have. Rodney Brooks for instance came up with this approach back in the early nineties. He called it the subsumption architecture, which is a fancy wording to "we have these nice basic procedures that work quite well in separate, let's mix them together and hope for the best". Yes, I am over-simplifying it. But it is not much more than that. Hey, just because something works it doesn't mean it has also to be complex. Take Roomba. It can't get any simpler than that. Or expensive and quite ineffective for what it costs.

So, pretty much everything in robotics is solved since the nineties. Right ? Well, no. The devil is in the details, or in this case what exactly is the combined behavior when you mash up things together. You see, roboticists have this fixation about concepts like system stability and, worse, predictability. Would you even consider to buy a Roomba if one of the possible behaviors was to steadily chip away the legs of your wife's beloved coffee table ? Not me.

The difficulty here is that to have a good idea of what the robot will do, and what it will not do, one has to put lot's of restrictions on what blocks can be used and how they are joined together. For sure, they are all mathematically beautiful restrictions, but restrictions nonetheless they are. And if you don't put enough, well then it comes up to a point where nothing can be said about the robot behavior.

What are the causes for the "false starts and few successes" of robotics ?

2010-11-18T20:02:00.001Z

In the previous post I commented on the bright, wonderful new future of mobile robotics that is being anticipated. Such cheerful prognostics commonly surface in the area, only to be proven wrong a few years latter down the road. I still remember Ronald Arkin giving an interview saying that in ten years time, robots would walk among us. This was said back in 1995. Or the famous objective of the RoboCup competition: "By mid-21st century, a team of fully autonomous humanoid robot soccer players shall win the soccer game, comply with the official rule of the FIFA, against the winner of the most recent World Cup." Which is now being recognized as "overly ambitious given the state of the art technology today". Although I must note that the last time the winners of the mid-size league played against a team of humans, they managed to score a goal. The Robocup dream is still far away from realization, but not something that it completely impossible.

The last hype in robotics, but certainly not least, was with their deployment in war theaters, Iraq and Afghanistan specially. Thanks to a better publicized than written book by Peter Singer, Wired for War, which suggested the increased importance of using robots in war. I don't know about that, but if it is true then why isn't the US winning the war in Iraq and Afghanistan ? Did the use of robots really tilted the field to the advantage of the US army ? I have my serious doubts, because the US army is not designed to fight the guerrilla wars it is facing, according to the better written than publicized Learning to Eat Soup With a Knife. Simply put, bigger and better guns don't win guerrilla wars and that has been demonstrated once too many times in history. In all of the three invasions the Napolean army, at the time the best fighting machine, made in Portugal at the start of the 19th century the guerrilla was fundamental in pushing them away. But I digress.

So, why isn't mobile robotics even close to what has been being promised over, at least, the last decade ? Certainly not due to the lack of quality manpower and resources. The cost of hardware as been decreasing for years, with the computation power increasing steadily. And there are many university institutes and research groups dedicated exclusively to robotics. So, what is it that is holding robotics back ?

A robot renaissance in underway

2010-11-18T11:22:00.000Z

"After decades of hype, false starts, and few successes, smart machines are finally ready for prime time. As part of its 2010 research, IFTF's Technology Horizons program has created the Robot Renaissance: the Future of Human-Machine Interaction Map to explore this new robotic future." says Lisa Munbach over at the Institute for the Future.

I don't share Lisa's confidence, because nobody has yet explained the motives behind the past "hype, false starts and few successes" in robotics. That being said, and knowing that most predictions about the future are likely to be wrong, the map [PDF] is still a good contribution for the discussion on the future of robotics.

Time is running scarce these days

2010-11-17T23:25:00.000Z

but I just to share with you my fascination about this little box of mistery: The Reverse Geocache™ Puzzle Box

It is as simple as it gets. Put inside a box: an arduino, a 16x2 lcd and a GPS. Oh, and I forget a button and batteries. Lock it such that it can only be opened when it is in the right location. Make it spew out clues when the button is pressed and give it to somebody special. Simple, special, brilliant.

Domestic robots

2010-10-21T16:12:00.000+01:00

The critter on the image is a robot able to "clean, sweep and do the laundry", developed at Tokyo University. Although the Roomba and other dust sweeping robots are known to most people, domestic robotics still has not delivered on its promisses. Like in so many other areas where robots are said to revolutionize in the next 10 years. As promissed every year for the last 30 or so years.

The thing that most people don't realize is that there two robots in this image, not just one. The washing machine is also a robot. It has sensors, typically temperature and water level, a controller, the washing program dial, and half a dozen actuators, the big rotor plus the water flow valves. This is ,by all accounts, a working robot. It is not able to load the clothes or to put them out to dry, but it is able to wash them without a problem. Notice that the key to the success of this washing robot is the succesful interaction with the environment. In particular, it relies on the environment to load the clothes into the washing drum and also to them them out.

This would not be a geek's blog

2010-10-10T09:26:00.003+01:00

if I had failed to notice this day.

Unwritten truths

2010-10-01T09:44:00.001+01:00

a big thanks to Scott Addams

Tongue twister

2010-09-28T14:38:00.000+01:00

Right now, some if not most of you are accusing me of playing with words. You are right, up to some point. After all a navigation device made by TomTom is not commonly considered a robot, although formally they are identical. The portable navigation gizmos sense the environment, do some processing on it and execute actions based on the processing results. Nothing else, nothing more than the operation of a Turing machine. It is fed symbols on a tape (sensing), changes the state (processing) and also the symbols on the tape (acting).

I would like now to point my illustrious audience (Google's spider robot and mummy) to this most curious fact. A TomTom gadget, a robot (humor me) without self-locomotion capabilities is able to reach an arbitrary point in the city park. But replace the human by a mobile base such as a Pioneer P3-AT, and you are lucky if it makes pass the rosebuds.

Notice that the problem is not the estimation of the position, it is the same in both cases. It is also not a problem of mechanic locomotion, since you can also build robots with humanoid kinematics. It is also not a problem in the thinking part, since it the TomTom gadget that in both cases computes the path and the actions to reach it. The only difference is that humans are able to ignore the GPS directions while the mechanic mobile bases are not. It is as if sometimes the instructions of the TomTom can/should be ignored.

A simple tought experiment

2010-09-23T23:05:00.000+01:00

Suppose that you have a robot that is positioned at an arbitrary location in Central Park (NY) and must reach an also arbitrary goal in the park. The robot is not able to sense the environment and is always executing the same action, going forward at a constant velocity. Finally, the robot is able to announce, in a textual manner, the goal location to surrounding humans. Do you think our little robot can ever reach the goal position in finite time ? Check the correct answer here.

Now, most of you will argue that it is not truly a robot since it cannot sense the environment. Well, suppose that the critter a) is equipped with a GPS which just happens to loose the position fix the instant the experiment starts and b) the program is such that in the absence of GPS information the robot moves forward at a constant velocity. Formally, both cases are identical.

Some amongst you, might cry foul. The robots, that is the objects sensing the environment and taking actions according to a list of instructions, are in fact the people in the street. It just happens that given the input symbol "smiling card box with written text", the people/robots react with the action "point smiling card box in correct direction". There is some research which suggests humans internally represent maps in Euclidean form, among other formats.

But ok, these are very reasonable objections. So now let the robot have an always working GPS receiver, it can announce verbally and visually the direction in which it wants to travel but, no free lunches, has no locomotion capabilities. Just to make it clear. The robot is able to sense the environment, compute a path to the goal position but has only one action: announce the desired direction of travel to surrounding humans. Can this second robot reach the desired goal ? Of course it can, here and here are many commercial examples of such robots.

Move along folks, nothing to see here

2010-09-21T14:10:00.000+01:00

Big deal. Robots are Turing machines. So what ? Everybody knows that.

An objection

2010-09-21T14:07:00.000+01:00

The claim that robots are formally equivalent to Turing machines is met with one serious objection. The only serious objection as far as I can see.

Probably the most common model for a robot is that of the agent, among others championed by Marvin Minsky. The whole operation of the robot is modeled by what is termed an "Sense-Think-Act" loop. The robot first senses the environment,
in the process updating the internal representation is has of the environment. It then computes the action it will execute next. The loop is closed by the environment reacting to the action of the robot. The new state of the environment is again sensed by the robot and the process repeats ad eternum.

So, my claim can be countered by pointing that I am reducing the robot to the Turing machine which does the thinking. However, notice that both the sensing and the acting parts of the loop can be accounted for, that is modeled, in the part of the loop that does the thinking. From a merely formal perspective, it does not matter how the model of the environment is updated or what is an action. This should not surprise anybody working in the field of Artificial Intelligence, where this formalization is a common practice. For instance, in a classic textbook of Reinforcement Learning, the formal model of agents is reduced to the controller.

Another, far less popular, approach to modeling mobile robots is treating the problem as the execution of distributed code on a graph, see the work by Peter Sapaty. Although it can be used to implement sense-think-act loops, it is more useful to model
distributed and/or decentralized systems such as multi-robot teams. In this case, it is clear that the system is a Turing machine since everything is code being executed on the different nodes of the graph.

The conclusion is then that a robot is formally equivalent to a Turing machine. For those more mathematically inclined, we can say that the property of being a Turing machine partitions the world in two. One half contains, among others, objects which are robots and the other half contains non-robot objects.

What is a robot ?

2010-09-01T11:02:00.002+01:00

A robot is a Turing machine. Or to put it in laymen terms, there is no formal difference between a robot and the device (computer, mobile phone, whatever) on which you are reading this. Whatever is true for a robot is also true for a generic computational machine, and vice-versa.

Oh ye of little competence

2010-08-25T19:50:00.000+01:00

So after realizing that the new shiny TV was a dud, I went to the store to replace it. Brought home a new one, which now is able to tune all the channels. But the evil gnomes of hardware were not yet finished with me. The new TV now reboots every 5 minutes or so. Didn't know that Samsung TV's came pre-installed with Windows.

Now, and before I go to the shop for the third time, let's pause for a second. The technological complexity of a personal mobile robot, say one that fetches beers from the fridge, is at least equal to these new high-tech TV's. At least computationally (code and hardware) there should not be many differences. Would you buy a personal robot if it meant having to go to the shop every couple of hours because of such silly flaws ? Or to put it in a different way. Would you spend your precious money and time in buying, what essentially is a new and innovative product, which only worked occasionally ?

A quick rant before moving on to more serious stuff

2010-08-24T21:50:00.000+01:00

I recently bought a Samsung Series 4 LED TV. Great image, superb sound, lots of inputs/outputs. Just one catch: it cannot tune any tv channels. None, niltch, niet, nada, rien de rien. After three painstaking hours, which included a call to Samsung's customer service, all I got was blank screen. Strangely this behemoth of technology is incapable of doing the basic. Something my old and cheap CRT TV does without a single problem. One definitely has to pause and question why the probably most technological advanced product in my house fails at such basic task. I mean, are Samsung engineers aware of the late 19th century technology called "bandpass filters" ?

In a rare display of Monty Python humor by the Samsung's engineers, this 7th wonder of TV technology always concludes the setup procedure with the message "Enjoy your TV". Ah.....

If Dante worked in robotics

2010-07-12T20:56:00.000+01:00

Inferno would just have only two circles: software and hardware.

I've just spent three exhausting weeks trying to setup mapping and localization on a small 8 meters by 12 meters area. What follows is a gruesome account of horrors, not advisable to sensitive souls.

First stop, the hardware circle of hell. In our lab there are four Pioneer P3-AT robots equipped with Sick LMS200 laser range finders. Or in laymen terms, about 60 K€ of equipment. Since these are some very expensive pieces of hardware, one would think it would be robust and efficient. At least that is what I would expect.

Problemo numero uno: the f***ing robots are made in the good old USA. That means imperial units for screws and stuff like that. That means spare parts easily available only in the other side of the pond. Here is a little example, two years ago one of the fuses in one of the robots blew. Shopped around all of Lisbon with no luck to find them. In the end, bought the european version and used a cutter to make it fit into the robot fuse holder. And why did the fuse blew ? Because one of the motors power MOSFET driver transistor blew. And it would be simple to repair if I could only unscrew the chassis cover with my metric hexagonal keys. Fortunately the power drill at the lab doesn't know the difference between a metric and a imperial head.... All in all, it took about 2 months to repair the bloody thing.

Problemo numero due: the Sick laser is a great piece of hardware. For the industrial environments it was designed for. Also, it is powered at 24V but the robots only have 12V available. Out pops an extra $100 for the DC/DC booster and say goodbye to a few watts of precious battery power. But the main problem is that, unlike theHokuyo lasers, there is no support from the company for drivers and shit. Which means we have to depend on engineering hacks. Thanks a lot SICK !!!

So you've managed past the hardware and must be thinking that by now, you can handle anything the universe throws at you. Think again, are you are entering the software circle of hell. The pioneers come with a good software library. It even supports the Sick laser. However I need mapping and localization capabilities which are not available by default. But hey, plenty of people have done this in the past and even published papers bragging about their fine accomplishments. Surely some of them have code available. Right ? Well, it is not that easy.

There is a nice site, OpenSlam, where the community leaves their code for others to use. The problem is that, well, the code is a bit academic. Or in other words, good enough to publish papers if you don't touch the magic constants in the code and know enough software jujitsu to compile it. However, compiling and running the software at OpenSlam leads to a dead end. It only gives you a map, usually as an occupancy grid, but I also need localization. My innocent self thought about writing a simple particle filter but hey, somebody certainly has done this already. Right ? Well, kind of. Again, most people in academia tends to write their code as they need it and then forget about it as soon as the paper is published.

So let's try a different approach, what about software frameworks ? What about the venerable lady CARMEN ? As it turns out, the lady is a tramp. CARMEN's was last updated two years ago and it is full of magical, mysterious, undocumented parameters and functions. Apparently their programming style guide is more of a wish list than an reality. After some Makefile jujistu I was able to compile the code only to discover that I could not build the map because the tool available, "vasco", kept on insisting in using the wrong units. After a couple of days conferencing with the resident CARMEN gurus, I admitted defeat and went out looking for other solutions.

Next on the line was MRPT, aka the Mobile Robot Programming Toolkit. Fancy name, fancy code, too bad it does not work. After spending one week digging through the poor documentation, trying to figure which options to use, the code would simply crash. SIGSEV was all she wrote.

Finally, I gave Player/Stage a try.And to my big surprise it worked. After four hours and a few posts to the users mailing list I had a working mapping and localization. Not state of the art, not excellent in performance but hey, at least it works. Finally a way out of hell !

The universe however, saw it differently, and pulled me back again to the hardware circle of hell. Out of the four robots, two have the WiFi connection out of order. Fixing it requires OpenSuse jujistu skills way above my belt rank. No problem, still have to robots, right ? Check again buster. One of the robots (remember it costs 15 K€) has a malfunction with the main power switch and only works if the charger cable is connected. Without it is an ugly, albeit expensive, pile of molecules. Ok, don't panic. You still have one robot working. Ah, but the evil hardware gnomes are not done with their mischiefs.

The Sick LMS200 has three leds: red, green and yellow. According to the manual, if the red is on and the yellow one is blinking it means there is a malfunction. According to the actual laser, it means you can still get data every two attempts. Good enough for me. It just means I have to have twice the work, but by now having something almost running counts as a big success.

Localization and identification

2010-06-24T11:21:00.000+01:00

The mantra for ambient intelligence: "localization is not identification". Essentially, ambient intelligence systems aim at enabling the environment to interact with the user. Yes, this requires localization capabilities. No, this is not sufficient. One thing is to be able to have interactivity occurring at some location, another is identifying what is at that location. Let me provide a few examples.

Suppose you have a robot, shaped like a garbage bin for example, that approaches people in a public park begging for garbage. The robot needs to be able to localize itself inside the park, but this is not sufficient. It needs also to identify the persons in the park. In another example, suppose that you keep on your mobile phone, a shopping list for the week. As you stroll down town or enter a supermarket, the mobile phone contacts the sellers computers to get the best deals for your shopping list. It then calls your attention as soon as you pass in front of the selected shops. Notice that in this example, interactivity requires both localization (where the person is) and identification (who the person is).

Where do we go now but nowhere ?*

2010-06-09T09:49:00.000+01:00

Ambient intelligence is pregnantfull of possibilities, so many that we feel bedazzled and paralyzed. So, where do we go now ?

An obvious choice are games, pervasive games that is. Image yourself a member of the secret organization Earth Defense Forces Against Inter-Dimensional Aliens, or EDFAIDA for short. You are walking peacefully down the street to your nine to five job, as all members EDFAIDA usually do. Suddenly, headquarters texts you on your mobile phone that a portal has just opened up at a nearby location, and vicious, slimy, ugly aliens are pouring in. You must catch them all and close the portal in ten minutes. Quickly, your mobile phone becomes an Inter-Dimensional Aliens Detection and Deletion Weapon, IDADDW, and the hunt is on. Thanks to the advanced augmented reality and positioning capabilities of your mobile phone, you can easily spot aliens and zap them. A few minutes latter, the area is alien free and the portal closed without the other people noticing you battle aliesn. These people never knew the danger they were saved from. You switch off the IDADDW and proceed to your job at an accounting firm. Back at the headquarters of EDFAIDA, the General increases your experience points and you get a promotion.

Ok, the acronyms can certainly be improved and the technology does not yet allow for this to be implemented, but you get the idea. What else can we do besides games ?

Tourism is another rather obvious choice. The simple minded content themselves with a GPS based automatic tourist guide. But there are much more possibilities. Suppose you arrive at Lisbon airport, incoming from Prague, for a few days of visit. At the tourism counter, strange maps get your attention. One invites you to visit Lisbon using the map of Prague. You wonder where on Lisbon, is your home street and the square close to where you work. Another map gives only geometric based routes. You discover Lisbon by moving along triangles, rectangles and hexagons. To follow the routes as accurate as possible, you enter private buldings, jump over fences, discover uncharted places. But the map you settle with, tells the legend of a princess and her lover, condemned by a jealous king to spend eternity separated on different hills of Lisbon. However, you can break the spell by assembling the Magic Rune Circle That Breaks The Spell (MRCTBTS), whose pieces are scattered through Lisbon. Collect them, and the two lovers become united.

Apart from my inability with acronyms, you see that tourist can be offered to travel a city using unconventional maps. This is a quite simple, but attractive idea since every maps leads to a different exploration and experience of the city. You invite people to leave the beaten track and experience the city in ways the locals probably don't even know.

So far we covered the obvious. What else ?

Borrowing from the idea of unconventional city maps, the same experience can be offered to shoppers in big department stores and shopping malls. You are the manager of one of such places and decide to offer customers an electronic shopping assistant. The customers input their shopping lists and go for a coffee. In the background, the assistant is busy looking for the best deals in the shops of the mall. When the customer returns it is given a list of shops which have the best deals on what she is looking for. The customer then picks up the assistant, and is guided towards the shops in the list. In another scenario, suppose you are planning on inviting some friends over on Saturday night. You go to your favorite supermarket, pick an electronic shopping assistant and ask it for suggestions on what to prepare. You decide for the Asian menu, hot and spicy, and the assistant guides you through the shopping activities.

Until now, I have only covered applications where people carry around devices to interact with the ambient. But, what if the ambient could interact with you ? Suppose you are a street performer, breathing fire while juggling chainsaws. The usual. Before starting you performance, you first approach a robotic public bench and ask it to follow you. You do this with a few more robotic benches until a small auditorium is in place. Finally, you approach a robotic garbage bin and tell to wander around the square, announcing your show to start in five minutes.

A bit sci-fi, but feasible. You can expand the idea with robotic lamp posts that follow people around and blink when they are touched. You can setup rendezvous with your friends, I will be the guy next to blinking lamp post.

* Yeah, it is a Nick Cave's song

First post

2010-05-22T12:46:00.000+01:00

Technology is changing the ways of the world. This is a truism, for sure, but there is more to it than what meets the eye. One very interesting trend is the increase in pervasive computation. Anyone, today, carries more computational power in his pocket than the computer that guided the Eagle to the Moon and back.

The raw computational power is used for mainly for communications and has decreased the distances between people. Friends on distant continents are now just a phone call away. More recently there have been some efforts in developing intelligent ambients, capable of interaction with humans. Through their phones mainly, but gamepads and netbooks have also been used.

The possibilities opened by the availability of cheap, pervasive computational hardware are limited only by the imagination. From role playing games in historical sites to intelligent shopping at supermarkets and even the materialization of social networks.

Coming from a background in mobile robotics, ambient intelligence is a very attractive area. Everything around us can now be transformed into a robot. Not that in the near future we will have flower pots that double as R2-D2, but they can certainly be made to think, communicate and move. An intelligent flower pot can move around, searching for ideal sunlight conditions and approach an intelligent water tap when it needs watering. And it does not stop here. Imagine public lamp posts that follow people around in a busy square. Or public benches which you can move around to build an instant amphitheater for street performances.

This may well seem like science fiction now, but there are certainly many open possibilities in ambient intelligence. This blog will be a diary of my journey exploring them.