3) Exercise on the topic: Motivated active decision making in autonomous robots.
Emotion & Threat Identifier
A simple diagram of Architecture for autonomous decision making a robot on an unknown planet 
For a Robot to explore an unknown planet in the search of living organism and minerals it must have various sensors and actuators, which are described below in detail. meanwhile, it has to escape any possible threats possessed by either terrain it is moving on or by the atmosphere it is in. however, if the robot gets an encounter with any living creatures, it must autonomously assess and identify threats from native species.
Robots/Rovers that are supposed to explore the unknown must have various sensors, for example, Multi-angle Cameras for photography, Laser sensors to measure lengths/distance, Temperature sensors, Radio Frequency Receivers, Light Sensors, Gyroscopic Sensors, Pizeo-electric Sensors, Compass, Atomic Clock, Spectrography etcetera. However, it is best for a robot to be equipped with RADAR and Motion Cameras, Richter Magnitude Scale and Anemometer to identify most possible threats.
From the above architecture, the input data from the various sensors, that include sensors allocated for threat detection goes directly to the “Decision processor”, where it categories input data either into emergency data or unknown data for now. Later the emergency data is sent to “Reflex System” to identify the level of threat that has been detected but it has a feedback mechanism which goes directly to “Control Unit” where it saves the experience, also if the threat requires immediate action it directly gives the appropriate command to “Motor System” to avoid the threat.
Now the unknown data previously collected and reflex data are sent to “Motivation System” where it examines the necessity of the data with respect the predetermined instructions in search of life as well as the instructions obtained by experience.
As “Control Unit” receives the data from both “Reflex System” and “Motivation System” where it process the data for further action and as well as storing the data as an experience separately but in sync, where the mood of the robot will be decided on based on the input and urges for survival and exploring.
Upon data Separation, in “Control Unit” based on behavior, instruction is moved to either “Explore Algorithm” or “Emotion or Threat Identifier”, basically robot draws the information required to use Actuators such as grippers, suction cups, drilling, and spectrography etcetera to identify minerals and water which are essentials for living creature survival, if instruction is taken into “Explore Algorithm”, else it would go for instincts and survival mode that are decided by “Emotion and Threat Identifier”, but the collective data is drawn by moto system and switched on for appropriate motion whether for exploring slowly or running away from threat.
Although with given predetermined motives and exercises exploring an unknown planet is always a risky process, various parameters are to be processed to know the unknown. In the above architecture, based on “Winner takes all” in “Reflex system”, “Voting based” in “Motivation System” and “Fixed Priorities” in “Explore Algorithm”.
1. Avila-Gracia, O. and Cañamero, L.D(2002). Comparing a Voting-Based Policy with Winner-Takes-All to Perform Action Selection in Motivation Agents. Department of computer science, University of Hertfordshire.
2. Breazeal, C.(2000)., Sociable Machines: Expressive Social Exchange Between Humans and Robots, Sc.D. dissertation, Department of Electrical Engineering and Computer Science, MIT.