A Novel Optimistic Local Path Planner: Agoraphilic Navigation Algorithm in Dynamic Environment

Round 1

Reviewer 1 Report

The paper presents a local path planner based on agoraphilic concept. The concept of agoraphilic is interesting and it calculates the obstacle-space to find the free space at every iteration. However, it seems that the algorithm continuously calculates the obstacle-space to find the free-space for navigation.

The presentation is good. Few grammatical errors were found.

Page no. 7: Difference between RR and RRR is not clear and LRR and LR.

Missing membership function for RRR in eq. 9.

Paper says there are eight firing rules. However, it is not clear why are there four indexing (that is i = 1,2,3,4) in eq. 12.

Sec. 2.8, line 3, spelling mistake for Weights.

It is necessary to conduct more experiments and simulations for the reliability of the proposed algorithm with different speeds of the moving obstacles and different locations for static obstacles.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors proposed a modified version of Agoraphilic algorithm to tackle the Dynamic Environments problem. It is compared with the baseline algorithms in real environment. The comments are given as below:

1.The literature review part is a bit poor. The state of art AI-based collision avoidance algorithm or bio-inspired collision avoidance algorithms should also be mentioned, for example, deep reinforcement learning algorithms.

2. The following two AI and bio-inspired collision avoidance algorithms are also worth for citation:

Accelerated sim-to-real deep reinforcement learning: Learning collision avoidance from human player. In 2021 IEEE/SICE International Symposium on System Integration (SII) (pp. 144-149). IEEE

Bio-Inspired Collision Avoidance in Swarm Systems via Deep Reinforcement Learning. IEEE Transactions on Vehicular Technology, 71(3), pp.2511-2526.

3.  Please also compare your work with the state of art method EGO-Planner [1].

[1] Ego-planner: An esdf-free gradient-based local planner for quadrotors. IEEE Robotics and Automation Letters. 2020 Dec 28;6(2):478-85

4. The author mentioned the proposed algorithm does not look for obstacles to avoid in the abstract, why do you still need model 2 (obstacle tracking module) and module 3 (dynamic obstacles position prediction) for your proposed algorithm in the section 2?  Please rephrase your word.

5, Please compare your collision avoidance algorithm processing time with the other state of art algorithms, as it is quite important to get a fast-computing time for avoiding instantly appeared fast-speed obstacles.

6, The proposed algorithm needs eight modules, please show the processing time of each module for the real-world experiment.

7, The experiment video is also worth to be shown, including the real world video, the rviz gui screen recorded video, etc.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for incorporating the suggestions

Author Response

Dear Sir/Mam,

We would like to thank you for taking your valuable time and effort to review the revised version of our manuscript titled “A Novel Optimistic Local Path Planner: Agoraphilic Navigation Algorithm in Dynamic Environment”. We sincerely appreciate your inspiring comments and accepting to sign the review report.  

Sincerely,

Hasitha Hewawasam