Simulation of human activities in smart spaces is becoming a necessary toolset to enable telehealth research that promises smart, humanly scalable and cost-effective solutions for aging, disabilities and independence. Human activities are complicated and contain a large number of varieties. Most of the existing research on activity simulations has relied on sensor events processing. Even though event-driven simulation is well established and practiced, it suffers from lack of scalability when applied to human activities. Hence, a new approach for scalable human simulation is urgently needed. We recently introduced a new simulation approach, which introduced a novel context based approach that leads to what we call a context-driven simulation approach. Our new approach is macroscopic and drives a set of multi sensor events (a context) rather as contrasted to the microscopic event-driven approach which is a single sensor event driven. In this paper, we present the detailed algorithms that implement the context driven approach. We present important concepts and simulation loop rules, followed by the algorithms. To validate our approach and algorithms, we simulate an environment and a scenario to examine the algorithms and the scalability of our approach in general. © 2013 IEEE.