That sounds like an excellent project. It's a lot of work for one person;
there are a lot of things to learn. How much time do you have - how many
weeks at how many days per week?
When you say you are a beginner, what do you know? What are you confident
with? Electronic design? Programming? Mechanical design? Actually,
physically building mechanics or electronics?
How far have you got? What have you built so far?
How big is the robot to be? A small, table-top robot (15cm) is a lot easier
than an outdoor robot (100cm). A lot of problems you'd meet with a big robot
can be safely ignored with a small one. For instance, you mention PWN (I
guess that's PWM control of the motors). With a small robot, you can get
probably away with simple on/off motor control. If you don't already know
how to make a PWM controller work then that could save you a week's work.
You say you're using phototransistors. Is that to spot the fire? It's more
usual to use a PIR sensor. PIR sensors are generally only used to measure
changes in temperature - not the absolute temperature. So I'd get the robot
to swing from side to side to sweep the sensor's view past the flame.
I realise that electronic components like PIR sensors can be hard to buy in
Pakistan but you'll be able to get PIR security sensors easily. They're
often easy to find in second-hand markets.