[[ Voting Sheet for Long-term Roadmap ]]
---------------------------------------------------------------------
name: Tomoichi Takahashi
Team: Hinomiyagura
---------------------------------------------------------------------
 We have to progress our RoboCup Rescue Simulation for contributing to
 the development of society. So, we want you to show your opinion for
 future RoboCup Rescue.

 Please answer the following questions.

 1. What do you hope next year?

 [a] We must keep and stabilize our current environment.
 [b] We must change our environment.
 Answer. [a or b]
---- b

 [What do those who selected "b" hope?                    ]

 Stability and challenge is chicken and egg argument.
 How about changing "stable" and "change" environments every other year :)
 So this year is a good chance, because
 2005-2006: Stable period:
            there is little change in environments.
            At Bremen, challenge items will be discussed.
 2006-2007: Change period:

 2. What do you hope next 5 years?

  At first, it is important to confirm the aims of RoboCup Rescue Simulations.
  a. platform of multi agent system,
  b. social agent simulations, modeling human/social behaviour,
  c. application to practical uses.

  a.) Yes, RSL provides a good platform of multi agent system.
      What are different from RoboCup Soccer Simulation ?
      We think the main feature of Rescue Simulation is social simulation.
      Namely, point b).

  b.) Agent approach provides a method of simulation to social science.
      Verifying simulation results requires to compare them with case data.
      This leads to point c).

  [Competition                                            ]
  a) size: map: the numbers of node & roads > 10,000
           agents: > 1,000
  b) change of ranking method
     unknown map and known map.
           background:
            Human rescue agents usually know well maps before disasters.
	    So, it is reasonable rescue agents also know GIS in advance.
	   new ranking:           
            trail 1: V1= simulation result under condition that agents
                         knows GIS data E1.
            disaster change E1 to E2 (network change about X %)
 
            trail 2: V2= simulation with the same code as V1 at E2,

            less (V1-V2), get higher point.
            notice: in case of V2 > V1, disqualify:
                    because, the team does not take advantage of E1 well.

  [Simulation                                             ]
  a) quality: collapse simulator occurs aftershocks.
  b) communication:    
            broadcast simulator:
              sometimes broadcast true data and other times rumors.
              


  [etc.                                                   ]