The complete proceedings of the AAMAS-08 Conference, held in Estoril, Portugal in May 2008, are now available FOR FREE from here.

The complete proceedings of AAMAS-07, held last year in Hawaii, are also available for free from here.

There is no catch to this, and it isn't a scam - you can access the complete proceedings for free without registering or similar. The proceedings are provided for the scientific community by the International Foundation for Autonomous Agents and Multi-Agent
Systems (IFAAMAS), a not-for-profit organisation whose activities involve organising the annual AAMAS conference. The aim of IFAAMAS is to make all future AAMAS conference proceedings similarly available for
free. The website of IFAAMAS is www.ifaamas.org

Founded in 2003, Kiva is focused on solving real problems in the supply chain. Founder and CEO Mick Mountz experienced the challenges of existing material handling systems firsthand while working at online grocer Webvan. The complexity of existing equipment and processes and the resulting high cost of filling orders ultimately was the downfall of Webvan. It occurred to Mick that there must be a better way to accomplish pick, pack and ship.

Mick then asked himself the simple question, "What if all the products in the warehouse could walk and talk on their own, couldn't they just come to me when I need to fill an order?" To pursue this idea, he sought the help of two experts in the area of complex multi-agent systems, Professors Peter Wurman and Raffaello D'Andrea, and together they began to develop the Kiva concept.

Today, Kiva applies the concepts of "distributed intelligence" to inventory management. Inspired by ant colonies capable of performing large and complex tasks with limited central control, the Kiva system allows inventory to organize itself, adapting to conditions as they change. The resulting solution combines store, move and sort functions into one simple system that can now deliver any item to any operator at any time.

If you are a new graduate, note that the are hiring.

1. Agents are problem solvers.
2. Agents are pro-active.
3. Agents are goal-oriented.
4. Agents are context-aware.
5. Agents are autonomous.

Perhaps. There are some who enjoy engaging in pseudo-philosophical discussion on whether or not a piece of code is pro-active, or goal-oriented, or context-aware, etc. In the end, however, that discussion completely overlooks the real contribution of multiagent research. Multiagent research provides the higher-level algorithms for organizing complex systems. Web services provide us with the plumbing needed to get one machine to talk to another and then multiagent algorithms tell us what, exactly, those services should be. The two operate at separate abstraction levels. There should not be any confusion between the two. It's like TCP/IP versus bittorrent.

More importantly (for us), is the fact that as web services (service oriented architectures (SOA)) increase in popularity there will be a corresponding increase in the demand for the interaction algorithms and protocols developed by multiagent researchers. The budding web services ecosystem will soon demand the creation of ever more sophisticated coordination mechanisms.

Let me reiterate my point with an example. One of the most commonly used multiagent examples is the trip-planning agent. This is an agent (call it a web application if you wish) which asks you where you want to travel, when, why, and keeps a profile of your preferences. It then goes off and makes all the required purchases for you: plane, car rental, hotel, tickets to a show, conference registration, etc. We can
easily envision airlines, hotels, and all the other service providers implementing a web-service backend to their website. However, this alone does not solve the problem. The agent must still figure out which specific things to buy for the user, how much to pay for them, and in what order to make the purchases (no refunds). Further, we know that the sellers themselves will also be changing prices to reflect demand. How do they make these decisions? Can two agents negotiate a deal that is better for both of them? Can we find more efficient (everyone does better) solutions by allowing more complex (multi-attribute, conditional, time sensitive) negotiations among agents? What added semantic markup do we need?

These are all very interesting questions which have not yet been fully answered. They are also the questions that multiagent systems research tries to answer. Not, “How do I make an online purchase from the Marriot hotel?”, but “What protocol and behaviors must all these agents implement in order to make trip-planning possible and incentive-compatible for all parties involved?”

I also noticed this quote:

But software, not hardware, may be the biggest challenge facing researchers working on swarms of robots, he [Alan Winfield] says: "Right now we just don't know how to design a system that produces complex overall behaviours from a group of simple agents."

Luckily for them, that is exactly the problem that multiagent researchers have been tackling for over a decade. Unluckily, it is a hard problem to solve, at least for some definitions of "complex ".

xaitment GmbH, one of the leading developers and service providers of artificial intelligence for the games and simulation industries, announced today that it has launched five new artificial intelligence (AI) modules ranging from standard to high-level AI functionality. The modules, which position the xaitEngine as one of the most flexible AI solutions on the market, will be shown at GDC 2008 in the German Pavilion (North Hall, Booth 6105).

The xaitEngine drives the predictable, and unpredictable, behaviors of computer-generated agents, and computer-operated applications and machines. It has primarily been used in interactive entertainment, though its applications are limitless - from the simulation of intelligent drivers in racing games, to resource management in manufacturing.

The xaitEngine handles all standard AI functionality, including pathfinding and simple AI behavior patterns such as movement. But while most other AI vendors stop there, the xaitEngine also handles more advanced AI. Such high-level AI can be found in the realistic interaction of the AI with its environment or the emotional intelligence of non-player characters (NPCs) in a game.

"Artificial intelligence has been offered in games for awhile," notes Dr. Andreas Gerber, CEO of xaitment. "But truly lifelike AI that offers emotional behaviors, autonomous actions and humanistic unpredictability, is something that almost no company has been able to provide, until now. What's more, we've been able to create our solution in a flexible manner that gives developers options so they are not locked into buying more than they need to enrich their own AI for their game."

As the marginal financial returns on graphical improvements decrease and the focus shifts over to playability, notice the popularity of the wii against the ps3 and xbox 360, we can expect to see more and more interest in using AI and multiagent techniques in games.

To me, the most interesting aspect of this development is that users will not want to battle infallibly intelligent opponents. That's boring. Instead the agents will need to incorporate emotional models, models of human-like non-rational behaviors (cf. behavioral Economics and Sociology for models), etc. They will also need to do this within a multi-player, thus multiagent, environment. In short, multiagent negotiation and decision-making techniques will find widespread adoption in the gaming domain.

The people at xaitment are betting on this, note their mission:

A spin-off of the world renowned German Research Center for Artificial Intelligence (DFKI), xaitment was founded in 2004 with the mission to create lifelike AI for games and simulations. Their mission led to the development of the xaitEngine, a highly customizable and highly modular multi-agent system that enables bots to learn from their mistakes, coordinate activities, compete with each other and achieve their goals with uncanny realism.

Two leaders may try to pull a swarm in opposite directions, and yet the swarm holds together. In Dr. Couzin?s model, the swarm was able to decide which leaders to follow.

?As we increased the difference of opinion between the informed individuals, the group would spontaneously come to a consensus and move in the direction chosen by the majority,? Dr. Couzin said. ?They can make these decisions without mathematics, without even recognizing each other or knowing that a decision has been made.?

You can get the original paper for all the mathematical details. I also found the following bit extremely amusing:

?Each cricket itself is a perfectly balanced source of nutrition,? Dr. Couzin said. ?So the crickets, every 17 seconds or so, try to attack other individuals. If you don?t move, you?re likely to be eaten.?

This collective movement causes the crickets to form vast swarms. ?All these crickets are on a forced march,? Dr. Couzin said. ?They?re trying to attack the crickets who are ahead, and they?re trying to avoid being eaten from behind.?

Now imagine them doing this in a circle.

Much of multiagent systems deals with the problem of (distributedly) computing equilibrium solutions for games, or calculating the proper incentives to that the agents will want to perform the required computation. Thus, much of multiagent research could be considered as algorithmic, or perhaps computational, game theory. There is a new book called Algorithmic Game Theory which collects short articles from the state of the art in this research area. David "oddhead" Pennock has more.

The thesis was titled "Keep the Adversary Guessing: Agent Security by Policy Randomization." Using highly refined equations and computer modeling, Paruchuri analyzed such situations as a security officer watching over a humanitarian relief camp for refugees and police officers patrolling a residential neighborhood that is prey to burglars. The formulas changed with varying numbers of players on each side, differing strategies and varying amounts of information that each side learned about adversaries.

The dilemma, Paruchuri wrote, is that police need "to commit to a security policy, while the adversaries may observe and exploit the policy committed to." He said he consulted with USC campus police on such topics as how to choose a patrol route.

Paruchuri's research shows how agent-based modeling used in conjunction with solid game-theory analysis make good things happen now.

Agent-based Systems are computer systems comprised on several autonomous, intelligent, communicating systems (or agents). This area is often referred to as Distributed Artificial Intelligence (DAI) or Multi-Agent Systems (MAS). Researchers in MAS build on techniques and inspirations from diverse areas such as game theory, economics, logic, and complex systems, in order to design efficient multi-agent systems. Practical applications of multi-agent systems range from automated trading in electronic markets, to robotic soccer, to distributed sensor networks, to grid computing, to intelligent workflow scheduling.

This school will be organized by the Autonomous and Adaptive Systems research group at the British University in Dubai, BUiD.

This article makes it sound like they have even grander plans:

"Successful peer-to-peer systems rely on designing rules that promote fair sharing of resources amongst users. Thus, they are both efficient and powerful computational and economic systems," David Parkes, John L. Loeb Associate Professor of the Natural Sciences at Harvard said. "Peer-to-peer has received a bad rap, however, because of its frequent association with illegal music or software downloads."

The researchers were inspired to use a version of the Tribler video sharing software as a model for an e-commerce system because of such flexibility, speed, and reliability.

"Our platform will provide fast downloads by ensuring sufficient uploads," explains Johan Pouwelse, an assistant professor at Delft University of Technology and the technical director of Tribler. "The next generation of peer-to-peer systems will provide an ideal marketplace not just for content, but for bandwidth in general."

The researchers envision an e-commerce model that connects users to a single global market, without any controlling company, network, or bank with bandwidth as the first true Internet "currency" for such a market.

They are proposing earn-and-spend market model, where the more a user uploads now and the higher the quality of the contributions, the more she would be able to download later and the faster the download speed.

But, wait, that's not all:

Another idea the researchers believe has enormous but untapped potential is the combination of social network technology with peer-to-peer systems.

I don't know if they will succeed, but I am sure these ideas will.

Ants seem to be popular lately. There is a new podcast on ants from Russ Roberts at econtalk where he interviews Deborah M Gordon on her book Ants at Work.

I first note how suboptimal the ants behavior can be. As she points out, foraging ants will actually walk over large piles of food in order to get to the original food source they had found in the day. In fact, they will continue to do this all day simply because it is not the forager's job to find food, it is the scout's and the scouts only go out earlier in the morning. Note also that it is the same ants who can play the role of either forager or scouts.
This inefficiency should not come as a surprise when we realize that ants evolved their fixed behaviors to survive in a very specific world, one where food evenly distributed over the landscape and large piles of food do not appear out of nowhere. Still, it should serve as a warning to anyone who thinks of building biologically-inspired multiagent systems. Similarly, anyone who thinks that a system is robust simply because it is "bio-inspired" is mistaken.

Russ and Deborah also had a discussion on the differences between the emergence in ants and in the economy. Namely, ants are hardwired to some very specific very simple behaviors while people are autonomous intelligent selfish agents. I note that most multiagent systems must lie somewhere between these two extremes. The simple behaviors of ants are rarely sufficient to get the job done while, at the other end, no one has yet built a generally intelligent agent. The question we face is: do I solve this bit of the problem with a simple clever behavior or do I have to increase the complexity of the agents?

Zhang [his student] is using a simple puzzle - the kind children can slide tiles around on to make a sequence of numbers [called the 8-puzzle]- as a starting point. She is looking at what it would take for robots to work together to solve the puzzle. For example, how would they move the tiles, how would they choose which tiles to move, how would they go about planning a way to solve the puzzle. As part of her doctoral work, she will list the gaps in scientific theory that would stop the robots from solving the puzzle together.

This is an interesting puzzle problem. First of all, lets realize that the centralized (single agent) version of this problem is a straightforward search problem which we can solve with A* or IDA* if we are concerned about memory. So, assuming we want to go distributed, we then have to decide what the robots can do. Are the robots like children all trying to move tiles in the same 8-puzzle? (which, in real life, always ends with a lot of screaming and crying). Or are they limited in some way?

One fun thing we could try is make each tile into a robot. Thus, each robotile wants to go to its destination and negotiates with other tiles to move on the empty space, if the space gets it closer to its goal. We could give them money and a simple bargaining strategy and see what ensues. I think this would work.

Another approach, more faithful to the children's story, would be to give each robot control of an area of the board and only allow it to move tiles that are in his area. When a robot has a tile near the border and wants to move into his neighbor's are they must enter negotiation. A tit-for-tat strategy would probably be effective in this scenario: if you let my tile in your country I'll let your tile in mine. I don't think this would work since they have no incentive to give the tiles to the neighbors that need them.

What happens when there is no leader? Starlings, bees, and ants manage just fine. In fact, they form staggeringly complicated societies, all without a Toscanini to conduct them into harmony. How? That?s our question this hour. We gaze down at the bottom-up logic of cities, Google, even our very own brains. Featured: author Steven Johnson, fire-flyologists John and Elizabeth Buck, biologist E.O. Wilson, Ant expert Debra Gordon, mathematician Steve Strogatz, economist James Surowiecki, and neurologists Oliver Sacks and Christof Koch.

The show also covers the wisdom of crowds and the neurological basis for consciousness (yes, you too are a multiagent system!). If you like the show, you might want to subscribe to the radiolab podcast feed.