Institutional Robotics and Norms in Multi-agent Systems

It is common in deontic logic to represent norms with the obligation, prohibition and permission  modalities. For instance, it is forbidden to steal.

In some of my publications I used a  rule engine to program the activation and deactivation of normative positions (a representation of these deontic modalities further restricted with arithmetical constraints) and their enforcement, namely some forbidden actions were sanctioned and other prevented.
However, in computer and robotic systems, deontic modalities are ambiguous since they do not state the type of enforcement the system will perform. For instance, what does it means that it is forbidden to steal? Will the system prevent all my attempts of stealing or will it punish them?

To overcome this situation, I proposed new types of rules with different enforcement meanings. That is, each rule has associated a different resulting behaviour when the system executes them.

I came up with these types of enforcement:

• Obligations
  1. Expecting: it is the softest type of obligation enforcement. Some agents are waiting for a certain obligation to be fulfilled and the agents that are expected to fulfill it are warned.
  2. Sanctioning: if the agent have not fulfilled a certain obligation yet, a punishment is applied.
  3. Forcing: it is the harshest type of obligation enforcement. Given some conditions and currently generated brute events, new brute events are generated in order to fulfill the obligation.
• Prohibitions
  1. Expecting: it is the softest type of prohibition enforcement. Some agents are waiting for the prohibition to do not be violated, and the agents that are expected to violate it are warned.
  2. Sanctioning: If an agent violates a prohibition, a punishment is applied. The difference resides in what types of events are sanctioned. Punishing brute events is harder than sanctioning institutional events.
  3. Ignoring: it is the harshest simple type of enforcement for prohibitions. However, the combination of all types of enforcement for prohibitions is even harder. Some brute events, as they cannot be prevented, are ignored in the process to translate them into institutional events.
• Permissions
  1. Expecting: some agents are waiting for some institutional events to be generated. The agents that are expected to generate them are informed.
  2. Allowing: brute events are allowed to be translated into institutional events that cause modifications in the context (or institution).
  3. Reward: If an agent generates a brute event that is translated into an institutional event, a reward is applied. The difference resides in what types of events are rewarded. Rewarding brute events is softer than reinforcing institutional events.

In language [1][2], I captured all this system behavior in the following way:

Expectations are captured with normative positions, i.e. representations of obligations, prohibitions and permissions. These are representation of deontic modalities in the form of atomic formula.

Allowing, sanctioning and rewarding are modeled through Event-Condition-Action (ECA) rules. Their meaning is that on the occurrence of some brute events, if conditions hold, then a list of actions are performed by the system. For instance, on waiving a hand if where are in an English auction then the system may consider that I’ve bid for the auctioned good, it may punish me because the waiving hands is forbidden, or it may reward me as the winner of the auctioned good.

Force rules behave like ECA-rules but also generating new forced brute events.
Then, ignore rules specify the brute events that must not be taken into account when evaluating ECA and force rules.

Institutional Robotics and Norms in Multi-agent systems by Andrés García-Camino is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Spain License.
Based on a work at blog.garcia-camino.es.

The goal of this post is to find better applications for Institutional Robotics. I present some of them and I will be updating the list. Please feel free to comment and increase this list. All your contributions will be credited.

• Defense Regulation: we cannot prevent military use of robots. However, its use might be regulated by neutral and trustworthy organisations as the ONU. For instance, robots are forbidden to kill civilians.
• Civil Regulation: in case robots were used by civilians in common-life tasks, they might be also regulated by each country in order to prevent misuses. For instance, prevent robots to steal, harm, kill, etc. following the orders of its owner.
• Self-Refereeing in RoboCup: adversarial teams of robots are currently being developed to play soccer, but the referees are still humans controlling robots remotely. A global view of the match, namely all sensor data from all robots, might be used to enforce the rules of the game transparently to robot reasoners [1, Future Work, New applications of norms].
  
  
  Aibo robots playing soccer
  An extract of my thesis, section New applications of norms:
  […]In this thesis, following the Electronic Institutions tradition we applied in our examples the regulation of Multi-agent systems (MAS) to e-commerce, mainly to auctions. However, we want to find and exploit new interesting applications of norms in MAS that may lead to enrich the current model. We envisage its application in robot soccer and autonomic networking.
  A question in the field of robot soccer that we find interesting is: how robots could autonomously play soccer without a referee? To solve this we envisage a software layer common to robots of any team. This layer would act as a middleware between the hardware layer, that deals with motors, rotors, sensors, etc., and the reasoning layer, that according with the information of the hardware layer generate an individual or team plan that is transformed in commands to the hardware layer in order to play. This middleware would follow the metaphor of institution but distributed in all robots to enforce some global rules. As a literary example, the reader may think of the hard-coding of Asimov’ three laws of robotics in the “brain” of every robot. However, the programming of the reasoning layer would be available of end-users, as human coaches in the case of robot soccer. This topic requires further exploration as, for instance, norms could vary in time possibly as result of the execution of protocols. For instance, soccer rules may change from season to season due to some recurrent unwanted behavior of players.[…]

Institutional Robotics and Norms in Multi-agent systems by Andrés García-Camino is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Spain License.
Based on a work at blog.garcia-camino.es.