1. 1 Energy-Efficient Mobile Robot Exploration Yongguo Mei, Yung-Hsiang Lu Purdue University ICRA06

2. 2 Outline Introduction Motivation Energy-Efficient Exploration Simulations Conclusion

3. 3 Introduction Explore an unknown area –Identify the locations of obstacles, objects, and free spaces using robots –Robots usually carry limited energy; thus energy conservation is important –issues Target selection motion planning

4. 4 Introduction Target selection –robot senses the environment while moving –robot accumulates the information from sensor data and constructs a map of the environment incrementally –At any moment, the robot needs to decide the next target to explore based upon the partial information Motion Planning –Plan a path to select target

5. 5 Motivation Frontier-Based Exploration [13] –To gain the most new information about the world, move to the boundary between open space and uncharted territory –All accessible space is contiguous, since a path must exist from the robot’s initial position to every accessible point –Every path that is partially in unknown territory will cross a frontier –If the robot does not incorporate the entire path at one time, then a new frontier will always exist further along the path

6. 6 Frontier –regions on the boundary between open space and unexplored space

7. 7

8. 8 Coordinated Multi-Robot Exploration [3] –select the next target based on the utilities and costs of the frontier cells –Utility is estimated based upon the size of new area that can be potentially covered at the frontier cell –Cost is estimated based on the shortest distance between the current location and this frontier cell

9. 9

10. 10

11. 11

12. 12 Energy-Efficient Exploration Environment –grid cell map –each cell is a 1 x 1 unit of square –each cell is either free or occupied by an obstacle –sensing range d s > 1 unit –robot's state : State(k) = ex: state(1) = –trajectory : State(0),….,State(k)

13. 13 Direction-based target selection –select the next target based on the relative directions of the frontier cells to the robot

14. 14 Target Selection Strategy 1.identify all the frontier cells that are within the current sensing region. If no such frontier cell exists, go to step (4). 2.lists all the frontier cells from step (1) in a clockwise order starting from robot's left direction 3.picks a frontier cell in the list that satisfies the following two conditions: (a) From list head to this frontier cell, any one cell and its next cell are neighbors. (b) The distance from the head to this cell is less than 0.7 ds 4.If no frontier cell is within the current sensing region, the algorithm picks the closest frontier cell outside the current sensing range as the next target. If there is no frontier cell at all, then all the accessible area has been explored and the exploration is completed.

15. 15

16. 16 Energy-Efficient Exploration Motion Planning –use Dijkstra algorithm to generate the minimum-energy paths and transform the grid map into a graph in a different way –incorporate the direction information –Each free cell in the grid map is transformed into 8 vertices –weight of one edge between two states is the energy needed for the robot to move from one state to the other state.

17. 17

18. 18 Simulations One unit of energy for traveling one unit of distance One stop takes an extra energy of 0.5. A turn of 45 0 takes 0.4 unit of energy. Turns of 90 0, 135 0, 180 0 take 0.6, 0.8 and 1 unit of energy d s = 10 units

19. 19 Simulations Energy-Efficient Motion Planning –path of length 36.28 vs. 39.21 (+8.1%) –consuming energy 49.28 vs. 45.41 (-7.5%) –3000 pairs : 0.7% longer, save 7.5% energy

20. 20 Target Selection and Robot Exploration –path of length 785 vs. 710 (-41.8%) –consuming energy 1222 vs. 1092 (-42.8%) –path of length 1207 vs. 1075 (-10.1%) –consuming energy 1390 vs. 1261 (-9.3%)

21. 21 Simulations

22. 22 Simulations

23. 23 Conclusion presented an approach for target selection and energy-efficient motion planning for robot exploration direction-based target selection reduce duplicate coverage, a common problem among greedy target selection methods. motion planning method considers the direction, stops and turnings, and can generate more energy-efficient paths. compared with a simple greedy method, our method can save up to 42% energy and 41% traveling distance in areas with structured obstacles.