Vision Based Mobile Robot Navigation System – Report And Analysis

State-of-the-art/Literature Review

The project proposal is based on “Vision based mobile robot navigation system”. The system plays a significant role and it is demanding for capability of the mobile robot. This navigation based mobile robot consists of different theories as well as technologies like odometry technique, vision system as well as ultrasonic mapping. Based on navigation application, it is mainly based on two areas such as global as well as local navigation (Siagian, Chang & Itti 2013). Global positioning system (GPS) and the Inertial Navigation System (INS) is used in different open areas. The local navigation is based on vision techniques, which is considered as an effective based for the closed range navigations. There are various types of applications within the indoor environment vehicles as well as mobile robots. Navigation with use of vision approach consists of various tasks such as target matching and identification of target (Ostafew, Schoellig & Barfoot 2014). There are two major factors which are affecting the accuracy of the navigation system such as optical sensors as well as processing of image. Into the application of navigation, mobile robot should interpret with the sensors data in order to extract the environmental information.

The paper is based on analyzing the selected project topic for the study. It provides with complete analysis on the topic with suggesting some of the research questions, aims as well as sub-goals for the selected topic. Methodology is summarized into this paper for identifying the methods and techniques used to conduct the study. Experimental set up is done to set the selected methods into the study. Results and outcomes are also provided to identify the possible outcomes of the project study. Apart from this, project planning is done to identify the possible project phrases of the project study. Project scheduling is done to identify all the project activities with its respective start as well as finish date.

Sharma, Chatterjee and Rakshit (2014) stated that different types of robots are used in the industrial applications. The existing application of the autonomous systems is one of the problems within the navigation system. If the working environment is unknown, then the navigation system becomes a high problem. The literature review of the project work presents with an approach to use of computer vision which is applying feature matching techniques. The features are such as circular marking, distance between the vertical lines used by fuzzy system. The vision system is applied in order to remove the information like distance of target as well as target orientations (Sales et al. 2014). The primary objective of the computer vision system is to segment the images in order to get useful information.

Huang et al. (2017) demonstrated that a search robot is navigating into the environment which has starting point, target object as well as various obstacles of random shape as well as sizes. It is seen that the starting point is known but the target point is not known. The robot is moving from the starting point to search for his target. The robot follows a continuous path and localizes itself in the environment.

Navigation: The exploration path

The strategies of navigation are based on the environment such as static and dynamic. These two types of categories are divided into known as well as unknown environments. Honegger et al. (2013) argued that before commencement of motion, the information is to be provided on location of obstacles. There are different types of navigation algorithms which is addressing the problem of robot navigation. All the navigation planning algorithms are assumed that the mobile robot consists of detailed knowledge of starting and targeting the locations (Schmid et al. 2014). Therefore, an optimal path is found among the two locations as well as it is a process to avoid the obstacles. Some of the algorithms are required environmental information and comprehensive map. Navigation algorithms are categorized into both global as well as local planning.

Global navigation planning: Chen et al. (2014) stated that these algorithms plan for the path of the robot from start to end towards the goal to search a graph which is representing a map of the global environment. This type of environmental graph is being constructed online or offline. The comprehensive map is loaded in the robot and then it is navigated algorithm is determining the optimal path before the robot is commenced its motion. Maohai et al. (2013) presented an optimal path planning algorithms which is suitable to search for the environmental workspaces in an image. The environmental view is being divided into discrete cells, and one of is the robot. This particular method is used to make use of camera at the fixed position.

Local navigation planning: Damaryam (2016) demonstrated that the local navigation planning is using the information of sensor within the commands which is controlling the motion of robot in each control cycle. It is not constructing a global map. This particular algorithm is employing to guide the robot into one straight path from start to the end point in order to target the location in dynamic environments. When the robot is navigating, it should avoid the possible obstacles that are in path and it keeps the information updated. Corke (2017) argued that the information are related to distance among the current and target location. The local navigation algorithms are not easy to build as well as it is optimal for the real time applications.

Leutenegger et al. (2015) stated that due to development into mobile robot navigation which is based on vision system are to be divided into two types such as indoor as well as outdoor navigation. Indoor navigation is categorized as map based and map building based. It is based on sequence of landmarks where the robot can able to detect the navigation. Outdoor navigation is based on sensor to construct the environment for the robots. Therefore, it can able to form an internal map for the purpose of navigation (Basaca-Preciado et al. 2014). Outdoor mobile robot navigation is categorized into two types such as structured as well as unstructured environments. Structured are needed landmarks in order to present the path of robot while unstructured environment is consisted of regular properties such as a vision system to extract the probable information of path.    

Strategies of navigation

The research questions for this particular project study are as follows:

• How should a self-navigation mobile robot control system design?
• How various types of sensors should integrate in the control system for search robot?
• How should minimum number of sensors attach on the body of robot for autonomous navigation?
• How should vision technology integrate with the robot technology for purpose of autonomous navigation?

The aims of this particular project study are development of mobile robotics system, which is proper for search robot which works simultaneously into indoor static environments. It is done to develop of a proper way to identify the location of robot orientation for efficient control of mechanisms. A proper method is selected to locate minimum number of sensors on the body of robot, which is perfect for autonomous navigation. The aim is to search for the optimal exploration path, which covers the complete environment of the search robot.

The goal of this project study is to produce a motion control system, which is able to employ camera as well as force sensors. It finds out a strategy to search for designed environment.

The robot control system enables the robot to search, find as well as relocate the target which is basically categorized into three parts. First it is to explore the task, which consists of finding proper exploration path (Jacobus et al. 2015). It covers the entire surroundings with insignificant sensing requirements and it constructs the path for the robots. Secondly, it is involved with finding proper image processing algorithms, which is suitable for detection of object and then implement and assess them into the robots. It also consists of rotating as well as then choosing of proper point of view field for the camera (Milford 2013). Third it is to relocate the task, which consists of gripper of robot in order to grasp detected wished object and shift it to assigned destination. The process of relocation involve with an approach both target as well as delivery site with use of visual tracking technique (Omari et al. 2015). In order to test the functionality of control system, there are two types of mobile robot such as software as well as electronic modules. First the design part of the robot is done such as sensor platform and then it is constructed to facilitate the robot to follow the map in the environment. Each of the functions of control system is being tested properly so as to receive the desired project outcomes.   

The experiments are conducted both online as well as offline mode. The environmental images are to be taken and it is processed in order to extract the features of object (Paolillo et al. 2017). It is probable to organize the environmental factors such as intensity of light as well as background colours. Therefore, the desired results and outcomes are obtained in this experiment. It is seen that the search robot is performing their task into dynamic environment while the object detection algorithm is evaluated into this condition (Nguyen et al. 2014). Therefore, the features of object are extracted from the images, which are captured into real time.

The first research question is based on design of self-navigation mobile robot control system. Basically, there are three skills required for the robot, which is supported by the optimal navigation system. The first skill is path-planning skill so that the robot can able to search his shortest path from the staring towards the targeted position. Existing navigation shows that the robot consists of ability to know proper location of the places. While in case of the search robot, he does not know the targeted position. The result is that the robot can navigate to investigate for the target. The robot can able to explore entire environment. The second skill is that the robot has capability to position itself within the environment. In that particular case, the robot must require to train and understand the environmental aspects. Finally, the robot has the skill to control the motion that is relied on motors, which move the robot from one place to another.

Vision based mobile robot navigation

In this project, with use of proposed navigation system, the robot starts their motion from initial location to the landmark. The main objective of the robot is to search, find as well as relocate the targeted object to the destination that is same as the starting point. In order to search for the research problem, the robot is mainly used to search for the indoor environment, which is definite by the external boundaries. The robot is navigating along the walls and it is considered all the project obstacles as the structure, which is similar to the wall. While the robot is moving, it continues to familiarize beside the wall. It keeps searching for the target location and the start point. When it is encountering a starting point, then it will regard as investigate process proficient and the target is inaccessible.

The following table shows the project schedule for this particular project plan:

WBS	Task Name	Duration	Start	Finish	Resource Names
0	Vision based mobile robot navigation system	90 days	Mon 6/12/17	Fri 10/13/17
1	Initiation phase	34 days	Mon 6/12/17	Thu 7/27/17
1.1	Identification of project objectives	2 days	Mon 6/12/17	Tue 6/13/17	Project leader, Project Manager
1.2	Business requirements	5 days	Mon 6/12/17	Fri 6/16/17	Project Manager, Senior management
1.3	Identification of business problem	3 days	Mon 6/19/17	Wed 6/21/17	Project leader, Project Manager
1.4	Documentation of proper responses	5 days	Thu 6/22/17	Wed 6/28/17	Project leader, Research Analyst
1.5	Business case	4 days	Thu 6/29/17	Tue 7/4/17	Research Analyst
1.6	Feasibility study	3 days	Wed 7/5/17	Fri 7/7/17	Senior management, Marketing Manager
1.7	Determination of recommended solution	5 days	Mon 7/10/17	Fri 7/14/17	Project customer, Project leader
1.8	Appointment of project manager	2 days	Mon 7/17/17	Tue 7/18/17	Senior management
1.9	Project deliverables	4 days	Wed 7/19/17	Mon 7/24/17	Product user group, Project leader, Project Manager
1.10	Approval of the project plan	3 days	Tue 7/25/17	Thu 7/27/17	Project Manager
1.11	Milestone 1: Completion of the initiation phase	0 days	Thu 7/27/17	Thu 7/27/17
2	Planning phase	23 days	Fri 7/28/17	Tue 8/29/17
2.1	Development of project solution	2 days	Fri 7/28/17	Mon 7/31/17	Line manager, Marketing Manager
2.2	Meeting with the project objectives	4 days	Tue 8/1/17	Fri 8/4/17	Project Manager
2.3	Identification of the project stakeholders	3 days	Mon 8/7/17	Wed 8/9/17	Project customer, Project leader
2.4	Outline the tasks and timeframes	1 day	Thu 8/10/17	Thu 8/10/17	Project leader, Project Manager, Project Sponsor
2.5	Coordination of project budget	2 days	Fri 8/11/17	Mon 8/14/17	Project Manager
2.6	Monitor and control of project cost	4 days	Tue 8/15/17	Fri 8/18/17	Marketing Manager, Product user group
2.7	Managing of risks	3 days	Mon 8/21/17	Wed 8/23/17	Marketing Manager, Product user group, Project Manager
2.8	Documentation of quality plan	2 days	Thu 8/24/17	Fri 8/25/17	Project Manager, Project Sponsor
2.9	Acceptance plan	2 days	Mon 8/28/17	Tue 8/29/17	Project tester, Research Analyst
2.10	Milestone 2: Completion of planning phase	0 days	Tue 8/29/17	Tue 8/29/17
3	Implementation phase	21 days	Wed 8/30/17	Wed 9/27/17
3.1	Performing the project work	3 days	Wed 8/30/17	Fri 9/1/17	Research Analyst, Senior management
3.2	Maintain control	5 days	Mon 9/4/17	Fri 9/8/17	Marketing Manager, Product user group
3.3	Communicate with the team member	2 days	Mon 9/11/17	Tue 9/12/17	Line manager, Project Manager
3.4	Monitor the project progress	1 day	Wed 9/13/17	Wed 9/13/17	Project leader, Project Manager
3.5	Carrying out the project task	3 days	Thu 9/14/17	Mon 9/18/17	Project Manager, Research Analyst
3.6	Performing the project meeting	5 days	Tue 9/19/17	Mon 9/25/17	Marketing Manager, Project Manager
3.7	Updating the project plan on regular basis	2 days	Tue 9/26/17	Wed 9/27/17	Project leader, Project Manager, Project Sponsor
3.8	Milestone 3: Completion of implementation phase	0 days	Wed 9/27/17	Wed 9/27/17
4	Closing phase	12 days	Thu 9/28/17	Fri 10/13/17
4.1	Releasing of the final deliverables	2 days	Thu 9/28/17	Fri 9/29/17	Project leader, Project Manager
4.2	Handing the project documentation	3 days	Mon 10/2/17	Wed 10/4/17	Product user group, Project customer
4.3	Termination of supplier contracts	2 days	Thu 10/5/17	Fri 10/6/17	Marketing Manager, Senior management
4.4	Releasing of project resources	1 day	Mon 10/9/17	Mon 10/9/17	Project Manager, Project tester
4.5	Communicating the closure plan	1 day	Tue 10/10/17	Tue 10/10/17	Marketing Manager, Project customer
4.6	Lessons learned document	3 days	Wed 10/11/17	Fri 10/13/17	Line manager, Project Manager
4.7	Milestone 4: Completion of closing phase	0 days	Fri 10/13/17	Fri 10/13/17

Following is the project milestone list:

Milestone activities	Expected Start Date
Milestone 1: Completion of the initiation phase	Thu 7/27/17
Milestone 2: Completion of planning phase	Tue 8/29/17
Milestone 3: Completion of implementation phase	Wed 9/27/17
Milestone 4: Completion of closing phase	Fri 10/13/17

Following is the work breakdown structure:

WBS	Task Name
0	Vision based mobile robot navigation system
1	Initiation phase
1.1	Identification of project objectives
1.2	Business requirements
1.3	Identification of business problem
1.4	Documentation of proper responses
1.5	Business case
1.6	Feasibility study
1.7	Determination of recommended solution
1.8	Appointment of project manager
1.9	Project deliverables
1.10	Approval of the project plan
1.11	Milestone 1: Completion of the initiation phase
2	Planning phase
2.1	Development of project solution
2.2	Meeting with the project objectives
2.3	Identification of the project stakeholders
2.4	Outline the tasks and timeframes
2.5	Coordination of project budget
2.6	Monitor and control of project cost
2.7	Managing of risks
2.8	Documentation of quality plan
2.9	Acceptance plan
2.10	Milestone 2: Completion of planning phase
3	Implementation phase
3.1	Performing the project work
3.2	Maintain control
3.3	Communicate with the team member
3.4	Monitor the project progress
3.5	Carrying out the project task
3.6	Performing the project meeting
3.7	Updating the project plan on regular basis
3.8	Milestone 3: Completion of implementation phase
4	Closing phase
4.1	Releasing of the final deliverables
4.2	Handing the project documentation
4.3	Termination of supplier contracts
4.4	Releasing of project resources
4.5	Communicating the closure plan
4.6	Lessons learned document
4.7	Milestone 4: Completion of closing phase

Conclusions

It is concluded that the project study is investigated on the search mobile robot and then find as well as relocate the target object within indoor environment. The navigation of robot is included to implement as well as simulate navigation algorithms for the purpose to validate the results from the experimental testing. It also consists of investigation of localization of robot such as simultaneous localization and then mapping. Detection of object is also included of vision sensor as it has capability to zoom and increase ability of robot in order to search for the larger areas. The construction of robot is becoming an interesting idea when there is building of larger scale robot and then it is searched for larger areas to relocate the objects. More and more sensor is being added into the robot in order to raise the ability of robot and explore into the environment. The robot intelligence is becoming important as there is requirement of autonomous system with use of artificial intelligent controller like the Fuzzy controller.

References

Basaca-Preciado, L.C., Sergiyenko, O.Y., Rodríguez-Quinonez, J.C., Garcia, X., Tyrsa, V.V., Rivas-Lopez, M., Hernandez-Balbuena, D., Mercorelli, P., Podrygalo, M., Gurko, A. & Tabakova, I., 2014. Optical 3D laser measurement system for navigation of autonomous mobile robot. Optics and Lasers in Engineering, 54, pp.159-169.

Chen, C., Chai, W., Zhang, Y. & Roth, H., 2014, May. A RGB and D vision aided multi-sensor system for indoor mobile robot and pedestrian seamless navigation. In Position, Location and Navigation Symposium-PLANS 2014, 2014 IEEE/ION (pp. 1020-1025). IEEE.

Corke, P., 2017. Introduction. In Robotics, Vision and Control (pp. 1-14). Springer International Publishing.

Damaryam, G.K., 2016. One Point Perspective Vanishing Point Estimation for Mobile Robot Vision Based Navigation System. International Journal of Research and Science, 5(2), pp.930-934.

Honegger, D., Meier, L., Tanskanen, P. & Pollefeys, M., 2013, May. An open source and open hardware embedded metric optical flow cmos camera for indoor and outdoor applications. In Robotics and Automation (ICRA), 2013 IEEE International Conference on (pp. 1736-1741). IEEE.

Huang, A.S., Bachrach, A., Henry, P., Krainin, M., Maturana, D., Fox, D. & Roy, N., 2017. Visual odometry and mapping for autonomous flight using an RGB-D camera. In Robotics Research (pp. 235-252). Springer International Publishing.

Jacobus, C.J., Beach, G.J. & Rowe, S., Cybernet Systems Corporation, 2015. Automated warehousing using robotic forklifts. U.S. Patent 8,965,561.

Leutenegger, S., Lynen, S., Bosse, M., Siegwart, R. & Furgale, P., 2015. Keyframe-based visual–inertial odometry using nonlinear optimization. The International Journal of Robotics Research, 34(3), pp.314-334.

Maohai, L., Han, W., Lining, S. & Zesu, C., 2013. Robust omnidirectional mobile robot topological navigation system using omnidirectional vision. Engineering applications of artificial intelligence, 26(8), pp.1942-1952.

Milford, M., 2013. Vision-based place recognition: how low can you go?. The International Journal of Robotics Research, 32(7), pp.766-789.

Nguyen, Q.H., Vu, H., Tran, T.H., Van Hamme, D., Veelaert, P., Philips, W. & Nguyen, Q.H., 2014, September. A visual slam system on mobile robot supporting localization services to visually impaired people. In European Conference on Computer Vision (pp. 716-729). Springer International Publishing.

Omari, S., Bloesch, M., Gohl, P. & Siegwart, R., 2015, May. Dense visual-inertial navigation system for mobile robots. In Robotics and Automation (ICRA), 2015 IEEE International Conference on (pp. 2634-2640). IEEE.

Ostafew, C.J., Schoellig, A.P. & Barfoot, T.D., 2014, May. Learning-based nonlinear model predictive control to improve vision-based mobile robot path-tracking in challenging outdoor environments. In Robotics and Automation (ICRA), 2014 IEEE International Conference on (pp. 4029-4036). IEEE.

Paolillo, A., Faragasso, A., Oriolo, G. & Vendittelli, M., 2017. Vision-based maze navigation for humanoid robots. Autonomous Robots, 41(2), pp.293-309.

Sales, D.O., Correa, D.O., Fernandes, L.C., Wolf, D.F. & Osório, F.S., 2014. Adaptive finite state machine based visual autonomous navigation system. Engineering Applications of Artificial Intelligence, 29, pp.152-162.

Schmid, K., Lutz, P., Tomi?, T., Mair, E. & Hirschmüller, H., 2014. Autonomous Vision?based Micro Air Vehicle for Indoor and Outdoor Navigation. Journal of Field Robotics, 31(4), pp.537-570.

Sharma, K.D., Chatterjee, A. & Rakshit, A., 2014. Harmony search-based hybrid stable adaptive fuzzy tracking controllers for vision-based mobile robot navigation. Machine Vision and Applications, 25(2), pp.405-419.

Siagian, C., Chang, C.K. & Itti, L., 2013, May. Mobile robot navigation system in outdoor pedestrian environment using vision-based road recognition. In Robotics and Automation (ICRA), 2013 IEEE International Conference on (pp. 564-571). IEEE.