Ph.D. student, Ning Xe, from the Department of Computer Science at the Tokyo Institute of Technology, recently sent the following letter to [email protected]:
Hi, Team members of Project Geppetto,
I checked the web page of Project Geppetto. This is really a fantastic project for crowd simulation. In our lab, we are conducting research on robot control based on a statistical machine learning method. Every year, we have a bunch of theoretical statistical machine learning papers published in journals and at conferences. Currently, we are looking forward to using some simulators to run and demonstrate our robot control algorithms for multi-agents. I would like to ask a favor as to whether we can plug in our algorithm to the core of Project Geppetto, for example, path finding or local behavior control.
Product Design Team Lead, Ian Nies, responded:
Project Geppetto currently can't be driven by an external path/behavior system but it’s a great idea and we will keep it under consideration. Thanks for your feedback!
I find Ning's suggestion quite curious. Is anyone else doing anything like this? Would any of you benefit from controlling your crowds from an external system? Technology previews are all about feedback. Project Geppetto is in its early stages. Now is the time to speak up. If you think this is a great idea, please let us know. If you think that the Project Geppetto team should spend its limited time on other aspects instead, we'd love to hear that as well. If the team shouldn't work on an API to allow the characters to be externally driven, what would you have them work on instead? You can provide feedback via:
If you have not yet but want to get started with Project Geppetto:
Inquiring minds are alive in the lab.