Robocon – A lifetime of learning and memories.

The ABU Asia-Pacific Robot Contest – Robocon is one of the most challenging robotics contest in the Asia-Pacific region. Every year a problem statement based on a certain country’s culture (the host) is released and requires the use of manual and automatic robots to complete the given task. This year, the contest was based on a tradition of Thailand known as Loy Krathog – “Lighting Happiness with Friendship”. A better picture of the contest can be obtained from the video itself:

Outline of the contest:

The Outline of the contest as obtained from this link is as follows:

Each team consists of no more than three robots: one manual and one or two automatic robots. The manual robot must complete the first task by picking up three Joss Stick Pots and placing them at Common Zone before performing other tasks. After that, the manual robot will bring a Candle Base and place it at Decoration Point located on Sala. The manual robot will collect Joss Sticks from the Common Zone to be used again during Krathong assembly.

The automatic robots will collect Krathong Petals and Flowers and place them at Preparation Points. The automatic robots will decorate Krathong by stacking one Krathong Petal and then one Flower on the Candle Base located on Sala. After completing this task, the manual robot will then place three Joss Sticks into the decorated Krathong. The automatic robots will carry the completed Krathong and drop it on River Surface of its own side. No part of any robots can touch or contact the River Surface.

Lastly, the automatic robots will bring and drop a Candle Light Flame on top of the Candle in the completed Krathong floating on the River Surface. No part of any robots can touch or contact the River Surface or the completed Krathong. The first team that drops the Candle Light Flame successfully is the winner of the match. This type of winning is called “Loy Krathong”.

If neither team achieves the “Loy Krathong” within 3 minutes, the winner shall be decided by the earning scores of the completed tasks.

Each match is contested by Red and Blue teams. A match lasts 3 minutes.

• A team must have one Manual robot and one or two automatic robots. The best way is to choose 1 Manual and 2 Automatic robots.
• The task was divided into 7 parts.

1. Manual-has to place the joss-stick pot at common zone.
2. Automatic1- has to place the petals and flowers (2 nos each) at preparation point.
3. Manual-has to place the candle base on decoration point.
4. Automatic2-has to stack the petals and flower on the candle base properly to make the Krathong.
5. Manual-has to place the 3joss-sticks on the stacked flower from joss-stick pot.
6. Automatic2-has to place the Krathong on the river surface.
7. Automatic1- has to place the candle flame on the Krathong to complete the game.

• This was the first time we participated in Robocon, so we did not have a clear picture of “how to” go for it. But since we were observing Robocon since last 2-3 years and also visited NIRMA institute to get more knowhow on it, it was a little relief.

Our Preparations:

Since Robocon problem statements are usually complex and challenging, very good preparation needs to be in place to even ‘just complete’ the tasks dictated by the problem statement. Good teams usually start 6 months ahead to make strategies and get their bots ready.

Due to delay in release of funds from the college, we had to start late this time. By 8^th January we got the funds and two of us rushed to Mumbai for parts. We started our work by Mid January. Without a guide it was really difficult to go on though.

We started with design of a test arena – one fourth part of the original one. Our strategy was to design one Manual and two automatic Robots to complete the tasks. Our task was divided into two major modules – mechanical & electronics.

 

Our Strategy:

Since we were a debutant team, we never expected to win. Our initial aim was to clear at least 2 levels. The focus was on the Manual (M) and First Automatic (A1) bots. We knew consistency is the key to win, so instead of focussing on the Second Automatic (A2) bot, we focussed on the full functioning and reliability of M and A1.

When we left the college, M, A1, A2 had all the mechanisms to complete their tasks.

Mechanisms and Design of the bots:

Our idea was to have huge precision, thus we had the following mechanisms in place:

• The Manual robot had the following components for its tasks:
  1. For basic locomotion – Motorised wheels on back and castor wheels on the front for free movement.
  2. On a horizontal two-slider base on which the whole arm mechanism was mounted.
  3. A vertical slider set was mounted on the movable base and it had all the arm mechanisms with it that gave them the desired height for functioning.
  4. The main arm was horizontal and extendible both sides and was capable of holding 3 joss-sticks pots at the same time.
  5. On one end of an arm was a mechanism to hold three joss-sticks from a joss-stick pot and on the other end was a claw mechanism to firmly hold and place the candle base.
  6. A switch box made of audio-amplifier switches to control the bot and its mechanisms.

• The First Automatic (A1) had the following components for its tasks:
  1. Same as M
  2. Vertical sliders on which the arm mechanisms were mounted.
  3. A composite arm for holding two flower parts at a time.
  4. Electronics Parts:
     1. Atmega 128 microcontrollers, Motor drivers, FRC converters, switch panels, etc.
     2. Industrial IR sensors (To sense the white lines), Bump Sensors (To sense the storage points), Magnetic Sensors (For restricting vertical slider movement)

• The Second Automatic (A2) had the following components for its tasks:
  1. Same as A1
  2. Same as A1
  3. An arm for holding one flower part at a time.
  4. A claw mechanism for holding the completed Krathong and dropping it on the river.
  5. Electronics Parts: Same as A1
• All the bots used:
  1. Rack and Pinion mechanism for sliding
  2. 150, 250 RPM square-gear motors
  3. 12V Battery power supply.
  4. Aluminium channels for construction

• The work progress: Our work went on till the end of February. We didn’t have the full arena for complete test but we had tested each bot individually 2 days before time. We thought precision was the only key to go ahead but actually speed of the bots also mattered. We couldn’t figure this out since we didn’t have the full arena to test.

Into the matches:

• The first match was with K J Somiya Institute of Technology, Mumbai. Since our manual bot was designed for precision, it was too slow to handle. It took almost 3 minutes to complete its first task. And since the total time was 3 minutes and the A1 bot was not allowed to run until M completes the task, A1 couldn’t be run. Our opponent was very weak though. They did not have any automatic bot or a complete manual bot design. We lost the match to them and got a serious wakeup call.
• To successfully run A1, we had to complete the first manual task in less than 1 minute. As a matter of fact, good teams like IIT Bombay, Nirma, could do it in not more than 10-15 seconds! We needed serious changes in M. And working hard overnight, we made the following changes:
  • The base slider mechanism was removed and the arm mechanism turned to front.
  • The arm was made fixed and sliders removed and instead of 3, it carried 2 joss stick pots. This is a good idea strategically since it helped to balance the bot. Only vertical sliders remain.
  • The 3 joss-stick holding mechanism was temporarily removed to balance the bot.
  • Guides were added below the arms for faster work.
• After these changes, the bot was reasonably fast. Now, instead of 3 minutes it took 40-50 seconds to complete its task. So sufficient time remained for A1 to complete its task.
• Since our bot was comparatively slow, the idea was to consistently score 70 points:
  • 18 from placing joss stick pots (M)
  • 40 from placing flower petals (A1)
  • 12 from placing candle base (M)

Full coordination was required between the bots and even the slightest failure would mean losing the match. You can very well guess about our consistency from the match videos. All of them seem to be alike!

• The idea of consistency was good. We beat Tezpur University with a heavy difference of 154-0.
  
  
  
• Some more corrections were made on the bot after problems arised:
  1. M was giving trouble with the arm mechanism to hold candle base, so we modified it again. It was pointing to left for dimensional reasons, but we managed to make it point forward for easy control by the operator.
  2. We also used relays in the main locomotion motor circuits since we found switch failures in test arena. But we didn’t have time to modify the whole switchbox.
  3. A1 didn’t hold the flower petals properly; the holding frames were re-made to put it in shape.
• The next match was with IIT Delhi, a strong contender. Key points about their bot:
  1. Use of Omni wheels (Wheels that can move both sides) on the base made very easy motion.
  2. Use of PS2 made control easy.
  3. But bots were big and unstable, i.e trembling arm mechanisms.
  4. Automatic bot (A1) was not working correctly.
• Our Automatic bot was quite stable and precise, it was slow though. So we could easily score the way we thought of and without A2 (the second automatic bot) we won the match.
  
  
  
• The fourth match was with VIT Pune, the runners-up of Robocon 2010. Key points about their bot:
  1. Manual bot was very fast.
  2. Was controlled by a keyboard, the operator seemed quite comfortable.
  3. The arm mechanisms were trembling and unstable.
  4. Automatic bot (A1) was not reliable.
• VIT Pune was defeated. It was an important match since it enabled us to enter the top 6. Two more teams were added based on their scores and the top 8 list was made.
  
  
  

• Finally the match we never anticipated, NIRMA University – 4 times Robocon winner and this time champions too. Key points about their bot:
  1. Small base & big aluminium wheels – for fast locomotion.
  2. Use of thin aluminium channels for light weight.
  3. Light-weight rubber sliders.
  4. Use of welding at many important spots instead of nut & bolt fixing thus giving more strength and stability.
  5. Simple bot structure – thus highly stable.
  6. Customized circuit boards – to keep things clean.
  7. Incredibly silent and fast motors – in-fact the fastest of all in Robocon 2011.
  8. Use of encoders – exact distance measurement, no need of sensing grids.
  9. Use of PS2 controller in manual bot – very easy control.
• NIRMA completed all its tasks in time and even assembled and extra Krathong. We didn’t score above 70 and thus lost it to NIRMA.
  

Drawbacks we never thought of:

1. Use of Thick aluminium channels, iron racks and steel sliders made the bot very heavy and it was slow. We should’ve used thinner channels, plastic/rubber racks and simple aluminium slider mechanisms.
2. Rack mechanism was too slow to handle, though it was precise. So sliding was slow. We should’ve used racks with larger teeth. We found that pulley mechanisms were faster and effective.
3. The base of the bot should be small, for easy locomotion of the bot. Our base was large and created heavy skidding problems with turns. An example of small base was NIRMA’s bot.
4. Use of Audio amplifier switches made it difficult for the motors to draw more current. Many a times the switches burnt out and caused failures. We should’ve used relays instead.
5. Secondly, due to the use of Audio Amplifier switches, the control was not easy either. Instead of our hand-made switchboard, a PS2 control is far smooth and comfortable to handle.

 

New ideas we saw in Robocon 2011:

Robocon was a rich learning experience. We discovered a lot of new mechanisms and problems that a bot can face any time. As such, here are a few new things we saw:

1. COE Pune used Pneumatic arms – they were very fast and effective.
2. Vertical lift was possible also with screw and pulley mechanisms. Pulley mechanism is really fast.
3. IIT Kanpur used Servo motors to turn the wheels for moving the bot in a different direction. So they didn’t have to turn the bot entirely.
4. Omni wheels were profusely used. They also avoided unnecessary turns.
5. Some colleges, for example IIT Bombay, used racks of microcontrollers to do parallel tasks.
6. Atmega 128 was in common use but PIC was also used by some.
7. Fibre arms and mechanisms were used to make the bot light. Ex: IIT Bombay bots.

Why NIRMA deserves to win? Not us? Not even IIT BOMBAY?

Of the many teams that we saw, the team from NIRMA UNIVERSITY, Ahmedabad is worth a lot of praise. And truly, NIRMA deserves to win every time. It is not just about “what they build” but also “how they build it” that matters. We got a chance to visit NIRMA University last year and got to know how they work it out.

The Mentor:

NIRMA has a highly dedicated Robocon team under the guidance of Mr. K.D Shah. Being firmly attached to Robocon, he’s the knowledge base of Robocon in NIRMA. From fixing small problems to even motivating and celebrating the win, we saw him as a true mentor and teammate in NIRMA. This is perhaps the most vital part of the team and the reason for their success.

Organization:

NIRMA had a large team. It consisted of 19 members (Maximum allowable) excluding the instructor and Alumni, who came as mentors. A typical Robocon team must consist of at least 15 members for even distribution of tasks. As we saw in the test arena, the team was very well organized. Each member was designated to do only a specified task and never interfered on others’ tasks. For example, there were separate members who took care of tools and others who used them. The whole thing was highly modular, which really gave perfection to it.

The quality of their bots:

When it comes to use of parts may it be motors or sensors, there was no comparision to NIRMA. As mentioned before, their bot was incredibly fast and soundless. That shows the quality of parts used. The Motors were imported from France and circuits were custom-made. NIRMA team never programmed their bot in the test arena. They ran simulations of their bot and tested for its mechanical stability even before the bot was built!
These are the three major aspects that distinguish NIRMA from every other Robocon team and make it a true winner.

Major constraints for Robotics in Burla:

1. Lack of Funds:

Robocon is all about very high precision and control. In a bot of dimensions 1m x 1m x 1.5m, even a 0.5 mm error will be fatal. So obviously we need quality parts for it. And that points to the big need – funds. As an example, the speed and precision of our motors were carefully calibrated in the program to compensate for their inaccuracy (Not a good idea). We could’ve use encoders instead, to get assured of high precision.

2. Availability of parts:

The College is very remotely located. We order parts from Mumbai that take weeks to reach Burla. So there is no chance to experiment. We’ve to be very sure of what we’re going to build. For example, use of heavy aluminium channels for the problem statement was not a good idea. The bots from NIRMA Institute were small, light and strong, and the secret was – experimenting with different materials. The solution is to have a huge variety and quantity of parts as it is with good Robocon teams.

3. Official support:

Robocon of course brought great laurels to the college, but it was never a smooth ride. More time was spent in official procedures and requesting for repeat mid-sems than building the bot. It would be very helpful if we have a permanent and dedicated instructor next time so that the knowledge is passed on to subsequent years easily and also the participants are free from lengthy official procedures.

About Funds:

One of the major challenges in Robocon 2011 was funds. We feel proud to mention that it was due to a major contribution of Rs. 85,000/- by Vedanta backed by one of our eminent alumnus Er. Asesh Padhi that enabled us to have the basic budget for the Robotics Society and Robocon. In addition, funds were also provided by the college towards the Robotics Society. As such, we were able to fulfil the basic requirements of Robotics Society and also availed parts for Robocon.

It is so possible that we can continue participating into Robocon 2012 only with the funds provided by college (The same amount allotted this year), but again, our performance won’t be satisfactory. We may give a stiff competition to the contenders but we can never expect to win. What we lacked this year is precision and quality. It can only be compensated by the use of quality parts. As an example, the main locomotion wheels used by NIRMA cost around Rs. 50,000/- each, while in our case it was Rs. 1,250/- each. That gives the reason for the silence and smooth functioning of their bots.

What we require now:

• What we basically plan is upgradation. Here is a rough sketch of the upgradable parts:
  1. Huge focus is on motors & motor drivers. (Rs. 20,000/- x 8 nos.)
  2. Encoders (For precise position)
  3. Light weight sliders
  4. Omni Wheels
  5. Other small circuits (development boards, PS2 controllers, etc.)
• We also plan to take training on ARM technology of robotics which will cost around Rs.25,000/- for 2 students.

Conclusion

We know that success is the only measure to know how and what we are working on. Till we taste it, there is no way to prove whether we are on the right path. With the kind of parts we had and the situation we were in this time, getting into the Quarter-Finals was a dream chase. But we gave the best performance possible to make our presence felt at the national level. Robotics is one such thing that incorporates knowledge from all fields of engineering and makes it highly challenging. In addition, it also boosts team building and management skills. Robocon 2011 was not just a landmark for Robotics in VSSUT but also gave VSSUT, high repute and recognition in national level. As such, robotics should be given top priority among college activities for academic and overall development of students.

Thanking you

Team Robocon 2011

Team Members:

Abhishek Routray (ETC, Final Yr.)
Sudhansu Sekhar Sahoo (Electrical, Final Yr.)
Shibashis Sahu (Electrical, Final Yr.)
Susant Mohapatra (Electrical, Final Yr.)
Ashis Kumar Sahoo (Comp. Sc., Final Yr.)
D. Ravi Rao (ETC, Final Yr.)

Barada Prasanna Beura (Comp. Sc., 3rd Yr.)
Rakesh Yadav (Electrical, 3rd Yr.)
Nitesh Kumar Das (Electrical, 3rd Yr.)

Sangram Keshari Das (Electrical, 1st Yr.)
Mentors/Alumnis
Biswajit Parida(2010, Mech.)
N. Seetharaman(2010, ETC)
Sarada Prasanna Nayak(2010, Electrical)