A thesis generally has the following chapters that include Introduction, Literature Review, Methodology, Analysis, Results and Conclusion.
The last chapter of the PhD thesis is the conclusion chapter and this is where any researcher can create a final impression of his work to the reader. Many researchers feel slothful while writing this chapter and may conclude it by just filling the pages with summary of previous chapters. In my view this is the chapter where the adjudicators observe i)how you visualize your results ii) have you identified any research gaps in your work iii) what limitations you had and v) whether you were really passionate about your work . The research scholars who are working in your domain will go through this chapter to see future research possibilities and limitations. A poor write up may create a bad image in the eyes of the adjudicator and the research scholars.
Conclusion section is the last chapter people read in your thesis, and therefore it’s what they leave remembering. You need to make sure they walk away thinking about your thesis just the way you want them to. The conclusion is something that a reader remembers the best. It often becomes hard for a reader to jot down all the important points of your thesis, in this case, a complete and meaningful conclusion take care to convey your study properly to the reader. A conclusion portrays the perfect picture of your thesis topic and it also delivers an idea whether the purpose of your thesis is achieved or not. It is a well-written chapter that summarizes and analyzes each of the thesis chapters.
- Revisit the work
- Highlight the big accomplishments.
- Recommendations for future work
In about one paragraph recap about what motivated you to take up this research work. Talk about the literature you have studied, discuss how your main findings from that literature are contradicted and justified by your research.Explain about the issues and challenges faced by the existing setup, recent technological and research advancements. Explain how you arrived at your problem statement and what your research aims and objectives were and what methodologies you used to tackle them.
Spend another paragraph explaining the highlights of your results. These are the main results you want the reader to remember after they put down the thesis, so ignore any small details.Try to keep the new findings in a greater relative importance and relates them to the issues in the world.
Your big accomplishments can be the data set you could able to gather or generate which can be used by researchers, government or Research agency support in the form of research grants, the improvements in results achieved for each objective, conversion of your idea into patent and product.
Recommendations for the Future work
A thesis looses its significance if it doesn’t usher new avenues for the future research. Most of the research scholars go through the conclusion to see if they find any unsolved problem for their research. You can make your thesis more meaningful by suggesting future possibilities from your work in conclusion section.
The future research section should be as succinct as possible. Each idea can be explained in one or two sentences by just giving the gist for each idea. Basic intention should be to motivate the researcher to take up the work in your area of research. This also helps you in getting better citations of your work in near future.
Points to ponder during conclusion writing
With learning the new ways of writing the conclusion, you should also remember some basic points to reach the conclusion.
Make use of your Creative writing skills
You know the analysis, literature, background and discussion chapters of a thesis are mostly dull and dry. However, a conclusion can be a platform for you to showcase your creative writing skills. Try to make your conclusion a two-way communication so that the readers can connect themselves to it.
You can write your own opinions about the results , analysis and future directions here. You can be authoritative here to share your wisdom and experience. Discuss your observations, place your insights on the implications of your study for future, and explicate about the scope of your work.
Accept the Limitations
Every research has some limitations. The limitations can be in the form of limited infrastructure,limited time , only consideration of either specific domain or specific environment for your research work. You can give justification for your limitations. Your research need not be perfect in all respects. There is always room for improvement. However, you need to accept that it is somehow limited. Your conclusion should appreciate that your thesis to has limitations and there may be a possible scope to add more content and research to it. But don’t forget that you have invested your valuable time in it, it is better not to be too negative about your work.
Dont Introduce Anything New
Writing too much and going out of the track is a genius trait, but you have to keep the very first line of this paragraph in your mind. The basic idea of a conclusion is, to sum up, things that you have already said before. There is no need to introduce new concepts and ideas in the conclusion. This will not only divert the reader’s mind from the main object but also create a chaos in his mind.
Few researchers repetitively write certain key words like “Chapter 1 discusses about” , “Chapter 2 discusses about “, “Our results are comparable to ” etc . You should avoid using such clichés and start your statements like ” As per the objective 1 the proposed work solves the problem of latency by introducing a new node ” , ” The proposed method is comparable to the industry standards set by…” . This will make your thesis more professional.
Finish with a rational thought
If you are proposing an algorithm and have proved that the algorithm is suitable only under certain circumstances or suitable only for specific set of domains, then make use of the conclusion to form a logical opinion without taking any side and back up your opinion by the evidence you have collected.
Be Confident of Your Work
Some researchers feel their work is not upto the mark or believe that they are not experts in the field of study. This they will showcase in conclusion by saying a “I may not be an expert”, “It is just a word of thought” or “I may be wrong, but”. This will dampen the whole spirit of the PhD work. So never make such silly statements.
Existing and potential applications of autonomous Robotic systems have one problem in common: Navigation. The navigational model actually functions as the interface between the sensory data and the control section and is responsible for the computation of a collision free path for a mobile Robot. There are numerous successful approaches to the Robot navigation task but Robotics research is still far from the vision of building Robots that behave much like humans and can truly cooperate and coexist with them. To be mobile, the Robot must be in a position to sense the environment, predict the motion of moving objects and find an optimal path to the target.
This work focuses on the core issues in navigation namely Vision, Navigational Environment and Motion Planning. These issues form the central part of sense-plan-act cycle of Robot motion planning. The navigation system developed here is applicable to any indoor environment consisting of rooms, corridors and doorways, which are accessible by a wheeled platform. This behavior fits mainly, but is not limited to personal and service Robots.
The proposed research work is funded by XYZ. Govt of India.
Performance of the object motion prediction algorithms developed in the work are verified by the real-life data sets available online. Real-life data has been gathered through benchmarked data sets from
i)INRIA Labs with data captured at INRIA Labs at Grenoble, France,
ii)Motion Capture Web group of Univ of S.California and
iii) CMU Graphics Lab dataset.
The data sets consist of different human motion patterns. These data sets include people walking alone, running, meeting with others, window shopping, entering and exiting shops.
The following sections discuss the major contributions made and the future research directions for Robot motion planning in 3D environment.
The work presented in this thesis has contributed three algorithms to the Robotics research at different levels of the sense-plan-act cycle.
A parallel multi-window stereo algorithm (SMW) for the depth map generation,on cluster computing setup is developed to overcome the slow response time of the existing multi-window algorithms. A detailed analysis of the implementation of the developed solution in terms of speedup, efficiency and suitability of the algorithm to parallel implementation is presented. The developed algorithm has achieved a speedup in the band from 1.48 to 1.68 and efficiency in the band from 0.74 to 0.84 on two node cluster. By adding one more node i.e. on three node cluster speedup varied in the band from 2.28 to 2.58 and efficiency in the band from 0.77 to 0.86. In most of the cases the efficiency of the algorithm is improved by the addition of a new node to the two node cluster. Based on the results it can be concluded that the performance of the algorithm initially improves with increase in number of processors but the observed improvement is less than the improvement expected by the Amdahl’s law in the cluster computing environment.
Ground suppression from the depth map is an important task in the Robot navigational applications. A novel methodology for ground surface suppression from the depth map using a Backpropagation neural network for partially known indoor environment is presented. The performance of the algorithm is tested in terms of the object reconstruction from ground extracted depth map and the histogram analysis. The algorithm has achieved an accuracy in the band from 86% to 98% within the training set and from 81% to 87% in the un-trained regions. The algorithm functions as a part of the depth map construction rather as an independent module, which saves considerable time of the 3D world reconstruction.
A parallel region growing algorithm to reconstruct the objects of interest from the ground suppressed depth map is presented. The navigation environment is segmented as Fuzzy distance layers and these Fuzzy layers are processed in parallel to reconstruct the objects of interest from the depth map. The method shows reduced response time when dealing with multiple objects in the navigational environment. The algorithm achieved a speedup in the band from 1.18 to 2.54 and efficiency in the band from 0.4 to 0.84. Variations in speedup are observed as the performance of the parallel object reconstruction is dependent on the number of objects in each Fuzzy layer. Increase in speedup is observed when objects are distributed evenly in the environment.
Future Research Directions
Future direction for research involves
- Adding more behaviors to the Neuro-Fuzzy based navigation algorithm.
- Developing a more powerful neural network architecture, that can fine tune as well as delete unnecessary Fuzzy rules in the Neuro-Fuzzy control architecture.
- Developing Neuro-Fuzzy control architecture for dynamic objects.
- Further, an emerging area of Robotic research “Swarm Robotics” will be considered for future work. Swarm Robotics is a new approach to the coordination of the multi-Robot systems which consists of large numbers of mostly simple physical Robots. It is supposed that a desired collective behavior emerges from the interactions between the Robots and interactions of Robots with the environment. This approach is emerged on the field of artificial swarm intelligence, as well as the biological studies of insects, ants and other fields in nature, where Swarm behavior occurs.
The work presented here looks more promising for real-life applications like developing Robotic wheel chair for the old aged and disabled, biped Robot as a hospital assistant and agro robot as a weed remover etc.