Exploring the use of Deep Reinforcement Learning to allocate tasks in Critical Adaptive Distributed Embedded Systems.