Spotronic & my first Mine Avoidance Robot

Written by Lim Siong Boon, last dated 11-Apr-09.

email:    contact->email_siongboon  

website: http://www.siongboon.com


 

Photos of SRG 1997

Robot Battlefield, laid with electronics mines. The mine will buzz and light up if it is being knock. Any robot that hit the mine will be out of the game.

 

 

Our senior's robot representing Singapore Polytechnic in SRG 1997. MAR uses 3 infrared sensor, one in the middle and two at the side.

 

Our senior's threat from Nanyang Polytechnic. Everybody were surprise by it performance in the 1997 competition. It has strong motor power and speed. This robot is able to complete the race before our senior MAR reaches their 2nd mine. The robot is able to stop quickly and negotiate the mine in a perfect curve in relatively short time, at least 5 times faster than MAR. It is no doubt a clear winner for Robot Battlefield 1997. No robot come close. It seems that they are using gear DC motor. The NYP robot have an array of sensor, sensing the mine from the top. A design that non of us expected. The design is a real genius, because they have break the impression that mines can only be sensed from the side. The position of the sensor is perfect, as mines sensing is very clear cut. This is unlike the side sensors position in MAR and the rest of the robots where sensing is more easily affected by the environment lightings when using infra-red. Looking into the components NYP use and the way the robot operate we guess that they are using PLC controller and a array of infra-red sensor. When the robot senses a defined pattern from the top-down sensor, it will negotiate the mine in a perfect curve close to the mine. Without side sensors to guide the robot, it is unlikely to negotiate the mine in such accuracy and speed. The only explanation for this behavior is that the curve movement profile is being pre-programmed into the PLC. Using PLC is the strength of their team, a group of mechantronic students.

Their design was great, and fit into the game. There is no rules at that time to restrict robot from using top down sensing. Since then, the SRG committee begin to seek new rule to deter new players to adopt such short-cut. The SRG 1998 committee begins to add on new odd shape obstacles prevent the use of such logic control robot. Example of new obstacles like a big/tall/non reflective dust bin in the middle of mine field, with low plank at the corner of the field. This new rules will ensure that the new generation of robot is intelligence enough to negotiate any type of obstacles and maneuver.

This robot from NYP becomes the ultimate aim in our project, to be better than this monster. Our team were looking forward to Robot Battlefield 1998, to fight with NYP robot. We will be following our senior's robot technology, using direct drive stepper motor (without gear). It was a real challenge.

 

Robot powered by a 9V battery

This is the robot, I most admire among all robot. If I remember correctly, this robot is from NUS. It is driven by 2 very small tamiya motor with small wheels power by a 9V Alkaline Battery. The whole robot is flat and light like a pizza box, and movement is very swift. Unfortunately there is a problem with their sensing and algorithm which hinder their robot to move between two mines. The robot just move forward and backward very swiftly, just like a mouse who is afraid to move through the mine field. If programmed correctly, this robot could be a winner. The problem might also be their choice of battery. The alkaline battery might not be suitable to run high current application. This may affect the power stability, which affects their digital logic circuitry. It will be extremely difficult to troubleshoot software if the bug is partly contribute by the hardware. In my heart, this NUS robot will always be special for it's performance in speed and energy efficient.

 

This is one of the robot using ultra sonic sensing. The sensor is enclose behind the casing. We are very aware that only this type of sensor at that time, can be mounted in this manner. We were trying very hard to analysis all possible threat for Robot Battlefield 1998.

 

Some other photos taken. It is a norm in SRG competition, that most IC chips are blank off, especially the important ones. This deter the competitor to learn more about their robotic system or to steal technological ideas. It is easy to learn about robot capability by observing their circuits, mechanical and sensing technology.

 

 

Spotronic Robot is still under development.

Motivate by my previous work in working on a mine avoidance vehicle, representing Singapore Polytechnic for the Singapore Robotic Games 1998. It is a new robotic event, and we are only the second batch of student competing in the event call "Robot Battlefield". It was a very exciting moment in my life that no amount of words can described that feeling. The competition barely a minute in the battle field determines our one and a half year worth of effort.

The happiest moment, our commitment have won us a Gold Award in the competition. The award is won by my team namely, project leader Chew Chuen Hee (hardware motor), Lee Chin Kwang (firmware sensor and decision control), Pang Sze Hsin (hardware electronics) and lastly me Lim Siong Boon (motor driver firmware).

The rules of Robot Battlefield can be found in the SRG 1998 website or a copy of the rules I saved from the website. A lot of  changes to the rule have been make for Robot Battlefield 99. Part of the changes are resulted from some frustrating incident that occur during our competition in 1998. I shall touch further on this in future.

 

 

The following are some of my design work regarding the robot sensor positioning. It has never been implemented because of the demand in writing motor driver software. I am suppose to handle all the software coding for the whole robot. In the end, the demand in writing the motor control algorithm has taken me the whole one and a half year of our project schedule. It was quite complex to me at that time because it is not easy to debug the mechanical performance in real time. Electrical problem may surface when the robot operate by itself on the floor. Switches get trigger due to mechanical vibration when the robot travels. These are the kind of unforeseen bugs makes our debugging task extremely difficult. The task of writing sensing algorithm is eventually taken over by my team mate Chin Kwang who is in charge of sensor design.

Mine Avoidance Vehicle project report. (13.4 Mb)

My motor driver software for motorola microcontroller MC68HC705C8CS (commonly known as 6805), master and slave operation for the left and right stepper motor. The file are in asm machine language. Although I have written a lot of comment in the source code, it is now meaningless to me. I hardly remember that they were my own codes. The concept is still clear in my mind, how the two stepper motors are operated. The robot uses 3 microcontroller because we have meet the speed limitation in motorola 6805. A single 6805 chip is too slow in supporting our target robot speed at 2 meter per second. We have suggested switching to a more powerful microcontroller but was rejected by our supervisor as he is afraid that the learning curve for a new controller will be long. In the end, we had all agreed to implement 3 microcontroller to control the robot. One main controller for sensor algorithm, one master and slave microcontroller for left and right motor. Initially I have great doubt that individual controller for each motor will work. There might be synchronizing problem which affect the robot from traveling straight. Eventually after the first prototype, it seems working very well and the problem is in fact negligible. There was nothing to worry about after all.

 

 

 

 

 

My team mate Sze Hsin, and Chin Kwang. We take some photos during SRG 1997.

 

 

Our team leader Chuen Hee (Mechanical, Electronics hardware), Chin Kwang (Sensor, Robot algorithm firmware), Sze Hsin (Electronics hardware, PCB), Siong Boon (Me, Motor driver firmware & software tools).

Too bad that these are all the photos I have on hand. Furthermore they are blur. If you have more photos to share for SRG 1998, 1997, 1999, please forward me your collection. I will be excited to receive these precious.

Pinky (left), Brain (right).

Pinky & the Brain, the name has a very symbolic meaning. To conquer the world, the dream of Pinky and the Brain, which is what we want our robot to achieve.

 

 

 

 

 

We have created two similar robot for the competition. It is the cutest name we come up with. I love it. One robot, we call it "Pinky", and the other one " Brain". Combine together, they are call "Pinky and the Brain".

The saddest thing is that I didn't keep any of the photos we had. I cannot forgive myself for not taking any pictures of our champion robot. It is one of the regret of my life for not taking any photos. The only traces of the robot we have created were our project report, certificates and the only website link on SRG 98.

Anyone of you out there, have document or picture for SRG 1998, please let me know. I am eager to get in contact with you. If only our competitor see this, I am sure they have our Pinky and the Brain photos taken.

Our developed robot is far superior from our senior. I forgot how we eventually agreed to join into this robotic game category. What we observe in our senior's demonstration, there are plenty of opportunities for improvement. Our senior's robot has limited sensor capability and their robot speed and movement is rather slow. For each object blocking their robot path, the robot execute a stop and do rotation, moving forward and rotate movement until the robot negotiated the obstacle "mine". This wasted a lot of time and speed as time is required to do acceleration and braking.

My job is to write a motor driver firmware to provide ease of control for our main software (sensor algorithm), at the same time control our stepper motor to speed and precision. To improve our robot speed in the competition, we cannot afford to stop the robot. The turning have to be made immediately when obstacle is detected. The motor driver allows full speed control and full degree of left, right turn options. Efficient acceleration and de-acceleration of both the left and right motor are taken care by the driver controller. This ease processing power from the sensing controller which is programmed by my team mate Chin Kwang.

I will write more of this topic in future.

 

written by Siong Boon, 29 Nov 2005

   

 

 

www.pic-control.com, Singapore Research & Development R&D

Singapore Customized, custom made Electronics Circuits & Kits

 

 

My Spotronic  

 

 

Here are some snapshot of the robot I am building. It looks a bit like a trash robot because I did not spent too much money for it. I come across the wheels in a shop and it got me a very long time before I decided to buy the wheel. The coupling between the wheel and motor shaft is custom made, and it cause me about S$120. I had been waiting to resolve this wheel issue until I started working in my first job, and eventually had enough saving to spent some on my robot.

I forgot to mount on another PCB circuit which is the motor and LCD driver. The PCB board shown on the photos consist of the logic circuits. They uses less power and the digital control is run by my favourite microcontroller from Microchip, PIC16F877A.

The robot is in-completed as some of the electronics module is under research phase.

1) Power regulator, DC to DC

2) Wireless FM Transceiver

Other module involved and does not need further improvement are as follows,

1) Microcontroller PIC16F877

2) DC Motor driver L6203

3) Serial communication interface

4) LCD control

I want to make a good robot. Although I had purchase a DC to DC module to supply my electronics, I insist on robot completion using minimum funding. The DC to DC research is completed but I had yet put onto my robot. The DC to DC cost half the price of the module I brought and it provides me with the experience for future higher current application.

The next module I lack of is the wireless transceiver, which I never successfully have it working until recent FM transceiver research. The previous attempt fail, until the discovery of implementing an encoder and decoder for the data transmission. Finally my dream of a remote control robot is on the right path. Parallel digital bits transmission is tested to be working fine. The next step will be to try out serial communication through the transceiver.

Will update on my robot again once I have completed my documentation.

 

written by Siong Boon, 29 Nov 2005

   

www.pic-control.com, Singapore Network Ethernet WiFi RS232 RS485 USB I/O Controller



 

email:    contact->email_siongboon 

website: http://www.siongboon.com


 

 

Keyword: remote control robot autonomous robotic Singapore Robotic Games