Spring Validation Demonstration
Location: Scott Hall 3rd floor, in area leading to Hamerschlag Hall
Equipment: WiFi router, Khepera IV robots with LRF modules (3), laptop
Objectives: For an indoor environment, visit all generated task nodes in less than 3 minutes per 70 m2 area with the optimal fleet size: 3 agents.
- Communication between the agents and server even in large environments.
- Localization in a large environment with a known map.
- Generating and assigning tasks to agents.
- Path planning and collision avoidance during search.
| Demonstration 1 | Fleet size of 3 agents |
| Demonstration 2 | Fleet size of 1 agent |
| Demonstration |
| Agents will be placed at their initial starting positions within the operating area.The user will give the agents an initial position estimate, along with a map of the environment.The agents will be assigned tasks using the Multiple Traveling Salesman Problem (MTSP) algorithm.Server will generate paths that are collision free for the agents to followThe agents will localize in the environment, and track the trajectory to complete the tasks.When an agent completes a task, it is reassigned to a new one. Steps 3-7 are then repeated until all tasks are completed.Once all tasks are completed, the search is terminated and the agents return to base. |
| Success Criteria | Validation Method |
| M.F.1. System shall localize robots with a known map of the environment. | Agent location and sensor data will be displayed on Rviz. |
| M.F.2. System shall dynamically allocate tasks to agents in the fleet for optimizing time required for the search. | Routing path will be plotted on screen.Search time will be recorded and compared between Demo 1 and 2. |
| M.F.3. System shall ensure simultaneous navigation of agents which have been allocated a task. | Agent paths will be displayed on Rviz. |
| M.P.2. System will have downtime of less than 5s for busy agents.M.P.4. System will update agent status for the user at 1Hz frequency. | Agent status will be displayed on Rviz. |
| M.P.5. System will handle at least 3 agents in the fleet. | There will be 3 robots in Demo 1. |