Undergraduate Projects
MAHVS - Multiple Autonomous Heterogenous Vehicle Simulation
Our client, Dr. Chris Stanton and our supervisor, Dr. John Taylor are researchers for The MARCS Institute for Brain, Behavior and Development at Western Sydney University, currently performing research on the matter of Human-Machine Interaction. The current version of MAHVS, created for such a research purpose is an air-traffic control simulation application run on PCs. The simulation will pair the user with an artificial intelligence partner which will make calculated decisions for the user, giving them a choice whether to trust the AI or not.
With the surge of the recent pandemic, it is difficult to conduct research in a local environment due to restrictions and social unease. The main objective of this project is to develop an improved version of MAHVS that is accessible by the web to perform large-scale experiments whilst also implementing a cloud-based server to store and collect information and data.
Solution Description
The new version of MAHVS, namely MAHVS2 will be run on a browser. It uses the Firebase as the database server for storing user IDs, trial parameters and the users’ unidentifiable trial data. A unique user ID will be provided to the user in the form of a unique URL in order to validate authentication for the use of MAHVS2 on a browser. When MAHVS2 starts, a new user will need to accept a consent form. MAHVS2 will then load a set of predefined trial parameters for that user from Firebase. The trial parameters include the number of aeroplanes, speed of aeroplanes, fuel of aeroplanes, etc.
The mission of users in MAHVS2 is to let as many aeroplanes as possible land on an airstrip safely without colliding with other aeroplanes. If MAHVS2 detects that an aeroplane is going to collide with another aeroplane, a prediction warning will display in red on that aeroplanes. Users will then decide whether they want to change the direction of the aeroplane to avoid the collision. However, every aeroplane has only limited fuel and it will explode after using up all its fuel. After the user finishes the trial, MAHVS2 will save the trial results under the user’s record on the Firebase server. The trial results include the number of aeroplanes crashed, the number of aeroplanes landed, the score of the trial and the set of trial parameters.
Developed By:
Anthony Tran
Chau Wong
George Zygas
Jiaqi Chen