Ant colony optimization (ACO) is a meta-heuristic based on colony of artificial ants which work cooperatively, building solutions by moving on the problem graph and by communicating through artificial pheromone trails mimicking real ants. One of the active research directions is the application of ACO algorithms to solve dynamic shortest path problems. Solving traffic jams is one such problem where the cost i.e. time to travel increases during rush hours resulting in tremendous strain on daily commuters and chaos. This paper describes a new approach-DSATJ (Dynamic System for Avoiding Traffic Jam) which aims at choosing an alternative optimum path to avoid traffic jam and then resuming that same path again when the traffic is regulated. The approach is inspired by variants of ACO algorithms. Traffic jam is detected through pheromone values on edges which are updated according to goodness of solution on the optimal tours only. Randomness is introduced in the probability function to ensure maximum exploration by ants. Experiments were carried out with the partial road map of North-West region of Delhi, India, to observe the performance of our approach.