I will be occasionally posting details of books that I have come across, which some may find interesting. This book gives a very good review of what is out there in the Crowd Simulation area. This includes different disciplines such as physics, sociology, computer science (as I mentioned in my first post), and the areas it's used in, such as the film and games industry, the academic research areas, planning, evacuation scenarios, and so on.
One of the authors, Daniel Thalmann, has quite a few publications in the Crowd simulation area, which can be found here As I mentioned in my first post that I will be exploring some of the above mentioned areas. This book also interestingly mentions that 'despite the apparent breadth of the crowd simulation research basis, interdisciplinary exchange of ideas is rare; researchers in one field are usually not very aware of works done in other fields'. I find this point very interesting, and therefore, brings to mind that work in various fields may often be duplicated, although I havent yet come across duplicate work, it is worth bearing in mind. I welcome such a book that would explore work that has been carried out in different disciplines.
The first few posts will be a catch up of what has been done so far. The first piece I worked on was a Netlogo to 3D Max implementation. This implementation is for a street level scale of a city. Although, 3D Max is used for crowd and particle simulations, in order to model these complex systems, an external package like Netlogo is used. Through the use of loose coupling with 3D Max, it allows us to easily visualise these systems.
The movie below was first posted on the Digital Urban blog. This was my first step to export a basic traffic model from Netlogo to 3D Max. The movements of the cars were first output into a file, the second step was to write a 3DS script to import these movements and animate them in 3D Max.
NetLogo to 3D Max - Proof of Concept from digitalurban on Vimeo. I have noticed quite a few comments asking how this import was done. I will be posting a pseudocode soon, and will try and give a summary for each step.
As this is my first post, let me detail a little about me, and about this blog. I am currently studying for a PhD in Crowd Simulation in the Centre of Advanced Spatial Analysis, University College London (UCL). I did an MSc in Vision, Imaging & Virtual Environments (VIVE) at UCL last year. I am in my 1st year of the PhD, and will be regularly updating this blog to include posts related to my work, work related to this area, and applications that I come across in the field. My main aim for the PhD will be to create a real time scalable crowd simulation, therefore, allowing users to look at, for example, crowds in the city of london at varying scales, in real time. This will be achieved using different architectures depending on the scale. The architectures I will be exploring will include disciplines in social science and computer science. The work will carried out on the Graphical Processing Unit (GPU), concentrating on general purpose programming, i.e., GPGPU. The reason I have chosen to use general purpose programming on the GPU, is because, there has been active research in this area over the last few years with the wide deployment of GPUs. The parallel processing of GPUs also gives us a distinct advantage to populate larger amounts of the city with agents, or a larger number of agents in a smaller area.
