Fritz Heckel

• AgentCore: Incomplete
• ThreadManager: Not started
• WorkerThread: Not started
• AgentController: Incomplete
• ActionModel: Incomplete
• PerceptualModel: Incomplete
• MemoryModel: Incomplete
• SubsumptionPolicy: Incomplete
• SubsumptionComponent: Not started
• Trigger: Testing
• Multitrigger: Not started
• Behavior: Incomplete
• SubsumptionBehavior: Not started
• CompoundBehavior: Not started
• Percept: Complete (Tentative)
• Effector:Complete (Tentative)
• FI3RSTCommunication: Assigned to Casey

Scientific:

Accomplishments:

• Designed configurable agent architecture based on subsumption
• Perceptual framework including “behavior parameters” which modify subsumption behaviors
• Conjunctions/Disjunctions of subsumption triggers
• Influence points to provide more efficient environment information than influence maps

Challenges:

• Further integration of agent properties / behavior parameters
• Consideration of cognitive theory in certain portions of BEHAVE
• High-level planning

Engineering:

Accomplishments:

• Built python prototype of BEHAVE
• Tested BEHAVE with BehaviorShop and FI3RST

Problems:

• Currently engineering the 1.0 version of BEHAVE. I am building skeletons of the important objects to flush out any issues that I haven’t anticipated in building the basic design.
• Apart from the communication and ssps portions, I believe the most effective way to split the work will ultimately be with the triggers and behaviors.

Here’s a list of good sub-tasks for behave which could be farmed out:

1. Network Communications: Use boost::asio sockets implementation with boost::serialization. Create the endpoints in C++, and wrap with boost::python as necessary. I think this will be simpler than converting the C++ objects into Python objects for use with pickle. The objects which need to be serialized and sent are: EffectorSet, Percept. FI3RST will send std::vectors of Percepts and receive std::vectors of EffectorSets. BEHAVE will send std::vectors of EffectorSets, and receive std::vectors of Percepts.
2. SSPS Graph Algorithms: Modify SSPS to use boost::graph, so that we can take advantage of the boost::graph algorithms.

Scientific:

Accomplishments:

• …

Challenges:

• Need to design!

Engineering:

Accomplishments:

• Added approaching information to agent percepts, so it can now be used in triggers
• Added speed of approach information to agent percepts; the accuracy is a bit bad, and I think this is related to the threading issues. Timestamps on the data coming out of FI3RST would probably also help.

Problems:

• I can try to make everything episodic; I have a couple of ideas to try out. I’m guessing something simply time-based will work. PerceptManager can calculate these.
• Agree with Dr. Y that we should probably go ahead and start using UDP for communication. Can try and add this to the python version now, or we can integrate it in 1.0. I’m concerned that we really just need to get working on 1.0 at this point.

I really think we should start work on 1.0, at least for BEHAVE. There are issues to address still, but I think we have to start work on the new version if we’re going to have it done or mostly done before GDC.

I persuaded the model to pose at a reasonable frame in the radio animation, yet, something isn’t quite right. See if you can figure out what it is…