Adam

ADAM-I


ADAM-I

Type: Robotic pet
Purpose: Adam was my first autonomous robot, I built him basically to see if I could design and build a "simple" robot. Although Adam was not as simple as I had thought he would be, I did however, gain much knowledge about robotics, and a greater respect for the people who build robots, now knowing how hard it is and how much work it takes to design and build a working robot from scratch.
Controller: The controller I used for Adam is an Atmel89C52 produced by Atmel corp.. www.atmel.com
Software Development Environment/Methods: I use SDCC C-Cross compiler developed by Sandeep Dutta.
Sensors: For object detection I originally had planned to use whiskers, but after some tries( none of which worked very well ) I decided that infrared would be easier to use for object detection. So Adam now has an infra red sensor from Reynold Electronics. I am planning to add three cadmium sulfide photo cells ( photo resistors ) for seeking, tracking, or avoiding light. I am planning to add two more photo resistors, and a LED for light, to the bottom of Adam so that he can also function as a line tracker, allowing him to follow a black line drawn on white paper.
Actuators: Adam does not have any arms or grabbers.
Height: Adam stands about twenty centimeters tall.
Width: Adam is ten centimeters wide.
Length: Adam is about twenty centimeters long.
Weight: Adam weighs 750g.
Power source: Adam operates on two different power supplies; The first is a four AA batteries running through a five-volt regulator for the microcontroller and the electronics. The second power supply consists of another four AA batteries for the two servos.

Construction history:

During my under graduate studies at College of Engineering, Trivandrum, I used to read a lot about walking robots and their control systems I even tried to build a mock leg to better understand the mechanics of the legs and how the robot walked but it was a total failure, designing precise gear systems proved to be too much for me. This was around April '96. After researching some more on walking robots, I decided that I was going to try to design and build my own eight-legged walking robot, but about two months into the designing process, I realized that I knew basically nothing about actually building or designing robots, especially walking ones! and the fabrication of mechanical parts was too much for me. The most important thing was getting the finance for these works. So I stopped developing robots for some time as I felt I was taxing my parents too much and I started concentrating on High frequency RF circuits which was purely electronics and I could manage RF circuits with my limited finance then. After two years in software industry I got too much involved in software and I wanted a break I thought of starting my interest in robotics again. One of the greatest impediment earlier was finance but with that was not that of an issue now. So I decided to make sure that I could at least design and build a simple robot before I tried to build a walking one. For the "simple" robot I was going to design and build, I decided on a small wheeled one (caterpillar wheels), and so Adam began.

The first thing that I did in making Adam was to construct the 89C52 microcontroller board. 89C52, which is the Atmel variant of the highly popular 8051 microcontroller, was supposed to be one of the easiest micrcontrollers to use, it proved to be rather difficult. Though I had previously worked with 8051, 89C52 with 4K internal flash Memory and 256 byte RAM was quite new to me. The on-board flash memory made my life quite easy with no additional RAM chip or EEPROMs I could download my code directly into its internal flash RAM. For writing into its flash RAM I made a 89C52 flash programmer the circuit diagram could be downloaded from here. One problem that I did not anticipate was that the microcontroller was not able to source much current through its ports and I had to add extra pull-up resistors this I found out very late, after integrating my servos and I had to spend some sleepless nights to find out the problem with my servos as the Pulse Width Modulated output was not found to be varying linearly as I had expected. Also there was lot of noise introduced by the dc motor for which I needed large fiter capacitors to reduce it.

The second thing I did in building Adam was to construct the infra red sensor that I got from Reynold Electronics the sensor is an integrated package with an infrared transmitter and a receiver. Initially I thought the sensor worked on the time difference between the transmitted and the reflected ray but I with the speed of light being in the range of 3*10^8 m/s the time difference is tooo small for any electronic circuit to detect. The working of the sensor is very interesting, it senses the angle between the transmitted and received infra-red ray. An object very near produces large angle while an object very far will give lesser angle. The sensor outputs a voltage which is proportional to the above angle. The variation was not linear but still I could live with that after all I need to find out if some object is in front of Adam or not. Though its data sheet specifies its detection range as from 10 cm to 80 cm I found it to be much lesser than that.

The third thing I did was making the motor control servo. The motor control servo used L293D motor driver chip from National Semiconductors. The motors were controlled through a PWM signals generated by 74LS193 4 bit asynchronous preset binary counter and some logic gates. The circuit for this is quite simple and straight forward. The microcontroller outputs an 8 bit value to control the speed of the motor. The MSB indicates the direction and the subsequent 7 bits specifies a motor speed form 0 to 127 scale. The data outputted by the micro controller is latched by an 74LS573 octal latch. This value is used to preset the counter whenever the base clock generated by a dual 555 timer (LM556) outputs a low value.

Finally I added a transreceiver circuit to Adam so that it can communicate with my PC but the circuit is yet to start working and I am still finding a lot of hurdles in make it fool proof. The most important thing is in making its transmitter and receiver shield against external noise generated by the motor.

Operational description: I have programmed Adam to avoid obstacles

I will describe a scenario to make it easier to explain... One fine obstacle detecting day, Adam is running forward when all-of-a-sudden, BAM! Adam's infra red sensor detects an obstacle. Adam, not wanting a conflict with the obstacle, reverses and then turns his right servo causing him to turn in place to the left, away from the obstacle. Detecting that no obstacle is in front of it, Adam resumes his journey forward.

Future Enhancements/Plans: From now on I will probably just enhance or replace Adam's program.
Status: Adam is partly finished, and as for the software Adam serves as a kind of "test bed" for when I have a "bright idea" for a new program.