Dhananjay Tiwari

Technical University of Munich, Germany – Dhananjay Tiwari

Hi, I am Dhananjay Tiwari, a dual degree student in the Mechanical Engineering department, going to be in my fourth year now. I have been pursuing a minor in Systems and Control  Engineering since my 3rd semester. I am a part of the Student Satellite Team and I work in Guidance, Navigation and Control of CubeSats. In my free time, I like to make sketches that usually have a realistic approach. I keep on learning new techniques in making portraits and also like to make various cities with my own thoughts in them. 

 I have been interested in core engineering since my first year. I used to find robotics very cool and so why I decided to choose systems and control engineering minor after reading its reviews. Later on, it turned out that there is much cool mathematics in it and robotics was very far. As time passed, I started realising that the systems and control is a lot more than just robotics; it is the mathematics of many physical systems around us. Now I am more interested in studying dynamics and controls of mechanical and aerospace structures, for example, movement of flexible arms, attitude control of rigid bodies, satellite control, etc. I have also been working with Prof. Ravi Banavar from systems and control engineering for more than a year. During the process, I have realised that whatever engineering inventions we see around us, it takes a good level of understanding of the theory behind it and a considerable amount of time. As I progressed in the project, whenever I felt demotivated, I looked back in time. My knowledge was consistently increasing, which kept me motivated independently of the results. It mostly involved understanding concepts through the courses (which I had taken according to my interests) or various reference papers and books, discussing with the professor and implementing them in simulations. I liked doing this, and because there were a lot of other things to learn and that too in some foreign university, where I would also get to see new people and their culture, I was very excited about going for a research internship. I started looking for various professors from various universities from Europe, US and Australia. My approach was to look for the right professors rather than just mailing too many professors. I used to visit the webpages of their research groups, their profiles and used to read the papers and works published by them thoroughly. I used to take suggestions about them from seniors and professors in controls at IITB. Some universities offer internship opportunities to international students, and they accept them through the internship portal. I came across the internship portal of the Technical University of Munich, Germany. I had to apply at the internship portal of TUM for which I also needed the approval of a supervisor from the TUM. Prof. Ravi Banavar recommended me to a professor at the TUM’s Chair of Automatic Controls, and I got approval from there within a couple of days via email :))    

Currently, my internship research is on the control of flexible robotic arms. In recent years, much attention has been given to modelling and control of flexible robotic manipulators. Conventional rigid robotic manipulators are often built to be heavy and bulky. However, they have certain disadvantages like high power consumption, low speed, high capacity actuators and low payload ratio. The robotic links are being made lighter to overcome these problems. The flexibility of light manipulators can not be ignored, mainly when they are operated at higher speeds. Hence, this study aims to accomplish a vibration-free point to point rotational manoeuvre, which is essential for fast and power-saving assembly in production, space applications, medical operations, etc. 

The current research to achieve the above goal is divided into two broad steps. First, deriving the equations of motion of the system, which in most of the cases is required by control algorithms to understand the system behaviour and act accordingly. The second step is to design the control algorithm itself, satisfying some constraints like limited vibrations. Currently, my work so far has been on understanding and simulating the dynamical behaviour of the manipulator, using a field of mathematical analysis called the calculus of variations. This has advantages over using Newton’s laws and Euler’s equations for flexible systems. Once the equations of motions are understood and validated through simulations, the next step would be to design an optimal control algorithm to move the manipulator from one position to another position in minimum time and reduced vibrations. The modular flexible link robot at the Chair of Automatic Control (TUM) consists of three revolute joints with Harmonic Drive gear motors, inverters and programmable logic microcontroller. The implementation of the frequency constrained optimal control design at the lab setup will serve as a proof of concept for the industrial exploitation of modern variational modelling and geometric control of mechanical systems (which I hope to do once we overcome this COVID crisis :)). 

Since the current work is hardware independent, the work from home experience is different from an actual visit in a sense that the interaction with the professors is less frequent, which otherwise, would have been daily. Although, the experience has been nice, as discussions happen on a regular basis, and there is freedom on both sides to present our thoughts with enough support. Well, there is much difference when we work at home, we tend to be lazy, and the efficiency of the work drops. To overcome this, I started making my notes (this is where I typed this article :p) of daily work updates and fixed up smaller goals too. This really helped me to monitor the progress of my work and cover-up if I am lagging behind. Easier said than done; it was really challenging while jumping over multiple references and finding a connection between to understand a particular concept; this required good enough patience. Implementing the concepts for simulation codes and debugging them is a different task in itself. There are always multiple errors spread in them over multiple files. To tackle this, I started documenting the debugging procedure, even the smallest of the results and errors, well this was something new which I have implemented, and it worked out well so far.

This overall process has involved studying a good detail of some advanced mathematics, and it became more exciting than I had expected when I started getting an intuition about it and relating it with the actual physical system. Although I am exploring a particular area in systems and controls through this project, I still know that the whole ocean of systems and controls is still left unexplored. I wouldn’t be able to explore all of that. I just have to find the right path and go deep down into it during my higher studies – this reminds me of the quote, “Your perception of your own knowledge is inversely proportional to your knowledge”. Getting an internship of one’s own choice is a difficult task but trying out something else is also okay if we are not sure of what we want. Remember that an internship is an opportunity to just explore something during this short period of a few months but we do not get expertise in that. 

I wish all the best to you, stay safe, healthy and happy :))

Thank you,


4 months ago