Method and system for operating a mobile robot

1. A method for operating a robot traveling in an environment, the method comprising:

the robot sensing the environment and thereby creating sensor data;

generating at least one probabilistic finding based on the sensor data, wherein the probabilistic finding is expressed as an object score and wherein the object score indicates a probability of detection of an object;

setting at least one detection threshold; and,

based on the at least one detection threshold, transforming the at least one probabilistic finding based on the sensor data to at least one discrete finding,

wherein the method is for operating a robot crossing a road,

wherein

the robot is configured to be controlled by at least one human operator when crossing the road, and wherein

setting the at least one detection threshold is based on a level of supervision by the at least one human operator when crossing the road.

2. The method according to claim 1 ,

wherein, in the transforming, a probabilistic finding below a detection threshold is transformed to a first finding and a probabilistic finding above the detection threshold is transformed to a second finding, the second finding being different from the first finding.

3. The method according to claim 2 ,

wherein the robot is configured to request assistance from at least one human operator,

wherein setting the at least one detection threshold is based on an availability of the at least one human operator, and

wherein the method further comprises:

the robot crossing the road without requesting assistance from the at least one human operator in response to the first finding; and

the robot requesting assistance from the at least one human operator in response to the second finding.

4. The method according to claim 1 , wherein the method is carried out by a data processing system comprising a data processing device external to the robot, wherein the method comprises:

a plurality of robots sensing the environment and thereby creating the sensor data;

transmitting the sensor data from the plurality of robots to the data processing device; and

the data processing device performing at least one detection algorithm utilizing the sensor data to generate the at least one probabilistic finding.

5. The method according to claim 4 , wherein

the at least one detection algorithm is a plurality of detection algorithms generating a plurality of probabilistic findings, and

wherein the method comprises for each detection algorithm:

setting a detection threshold combination comprising a detection threshold for each probabilistic finding,

utilizing the detection threshold combination to generate a discrete finding for each probabilistic finding, and

testing a validity for the detection threshold combination.

6. The method according to claim 5 ,

wherein testing the validity for the detection threshold combination comprises testing an individual finding validity for each discrete finding,

wherein the individual finding validity results in either one of a true positive finding, a true negative finding, a false positive finding, and a false negative finding,

wherein the detection threshold combination thus comprises a true positive rate, a true negative rate, a false positive rate, and a false negative rate.

7. The method according to claim 5 , wherein the method comprises:

setting a plurality of detection threshold combinations, each detection threshold combination comprising a detection threshold for each probabilistic finding,

utilizing each of the detection threshold combinations to generate a discrete finding for each probabilistic finding,

testing a validity for each of the detection threshold combinations.

8. The method according to claim 7 ,

wherein, for each detection threshold combination, testing the validity comprises testing an individual finding validity for each discrete finding,

wherein the individual finding validity results in either one of a true positive finding, a true negative finding, a false positive finding, and a false negative finding,

wherein, for each detection threshold combination, the detection threshold combination thus comprises a true positive rate, a true negative rate, a false positive rate, and a false negative rate.

9. The method according to claim 8 , wherein the method comprises:

setting a false positive rate, and

finding the detection threshold combination having a false positive rate not exceeding the set false positive rate and having the highest true positive rate of the plurality of detection threshold combinations.

10. The method according to claim 4 , wherein the method further comprises:

processing the sensor data to generate processed sensor data, wherein the at least one detection algorithm utilizes the processed sensor data to generate the at least one probabilistic finding.

11. The method according to claim 10 , wherein a ratio of data size between the sensor data and the processed sensor data is greater than 10.

12. A system for carrying out the method according to claim 1 , the system comprising a data processing system and a robot.

13. The method according to claim 1 , wherein the at least one detection threshold is based on a visibility in the environment.

14. The method according to claim 1 , wherein the at least one detection threshold is based on an amount of traffic.

15. The method according to claim 1 , wherein the at least one detection threshold is based on a type of the road.

16. The method according to claim 1 , wherein the at least one detection threshold is based on a distance traveled when crossing the road.

17. The method according to claim 1 , wherein the at least one detection threshold is based on a movement status of the robot.

18. A method for operating a robot traveling in an environment, the method comprising:

the robot sensing the environment and thereby creating sensor data;

generating at least one probabilistic finding based on the sensor data, wherein the probabilistic finding is expressed as an object score and wherein the object score indicates a probability of detection of an object;

setting at least one detection threshold; and,

based on the at least one detection threshold, transforming the at least one probabilistic finding based on the sensor data to at least one discrete finding,

wherein the method is for operating a robot crossing a road, and

wherein setting the at least one detection threshold is based on whether the road crossing is performed automatically or operator-based.

19. The method according to claim 18 , wherein the at least one detection threshold is based on a type of the road.

20. The method according to claim 18 , wherein the at least one detection threshold is based on a distance traveled when crossing the road.

21. The method according to claim 18 , wherein the at least one detection threshold is based on a movement status of the robot.

22. The method according to claim 18 , wherein

the robot is configured to be controlled by at least one human operator when crossing the road, and wherein

setting the at least one detection threshold is based on a level of supervision by the at least one human operator when crossing the road.

23. The method according to claim 18 ,

wherein, in the transforming, a probabilistic finding below a detection threshold is transformed to a first finding and a probabilistic finding above the detection threshold is transformed to a second finding, the second finding being different from the first finding.

24. The method according to claim 23 ,

wherein the robot is configured to request assistance from at least one human operator,

wherein setting the at least one detection threshold is based on an availability of the at least one human operator, and

wherein the method further comprises:

the robot crossing the road without requesting assistance from the at least one human operator in response to the first finding; and

the robot requesting assistance from the at least one human operator in response to the second finding.

25. The method according to claim 18 , wherein the method is carried out by a data processing system comprising a data processing device external to the robot, wherein the method comprises:

a plurality of robots sensing the environment and thereby creating the sensor data;

transmitting the sensor data from the plurality of robots to the data processing device; and

the data processing device performing at least one detection algorithm utilizing the sensor data to generate the at least one probabilistic finding.

26. The method according to claim 25 , wherein

the at least one detection algorithm is a plurality of detection algorithms generating a plurality of probabilistic findings, and

wherein the method comprises for each detection algorithm:

setting a detection threshold combination comprising a detection threshold for each probabilistic finding,

utilizing the detection threshold combination to generate a discrete finding for each probabilistic finding, and

testing a validity for the detection threshold combination.

27. The method according to claim 26 ,

wherein testing the validity for the detection threshold combination comprises testing an individual finding validity for each discrete finding,

wherein the individual finding validity results in either one of a true positive finding, a true negative finding, a false positive finding, and a false negative finding,

wherein the detection threshold combination thus comprises a true positive rate, a true negative rate, a false positive rate, and a false negative rate.

28. The method according to claim 26 , wherein the method comprises:

setting a plurality of detection threshold combinations, each detection threshold combination comprising a detection threshold for each probabilistic finding,

utilizing each of the detection threshold combinations to generate a discrete finding for each probabilistic finding,

testing a validity for each of the detection threshold combinations.

29. The method according to claim 28 ,

wherein, for each detection threshold combination, testing the validity comprises testing an individual finding validity for each discrete finding,

wherein the individual finding validity results in either one of a true positive finding, a true negative finding, a false positive finding, and a false negative finding,

wherein, for each detection threshold combination, the detection threshold combination thus comprises a true positive rate, a true negative rate, a false positive rate, and a false negative rate.

30. The method according to claim 29 , wherein the method comprises:

setting a false positive rate, and

finding the detection threshold combination having a false positive rate not exceeding the set false positive rate and having the highest true positive rate of the plurality of detection threshold combinations.

31. The method according to claim 25 , wherein the method further comprises:

processing the sensor data to generate processed sensor data, wherein the at least one detection algorithm utilizes the processed sensor data to generate the at least one probabilistic finding.

32. The method according to claim 31 , wherein a ratio of data size between the sensor data and the processed sensor data is greater than 10.

33. A system for carrying out the method according to claim 18 , the system comprising a data processing system and a robot.

34. The method according to claim 18 , wherein the at least one detection threshold is based on a visibility in the environment.

35. The method according to claim 18 , wherein the at least one detection threshold is based on an amount of traffic.