Abstract-Machine Type Communications (MTC) collectively refers to the exchange of data among devices that operate without human intervention. A significant number of such devices are currently served by cellular networks, and that number is expected to grow in the near future. However, cellular networks, including current fourth generation LTE networks, face several challenges when it comes to handling MTC traffic as they were primarily designed for Human Type Communications (HTC) which have very different traffic patterns. In this paper we focus on periodic MTC devices, such as sensors and meters, which cause significant signaling load and increased collisions over standard LTE networks. We propose ASPIS, a holistic mechanism designed to overcome these problems. ASPIS reduces the signaling load by partly preserving a device's connection to the network in conjunction with a new Random Access process and efficient support for short message transmissions. In addition, it uses a proactive preamble split scheme to alleviate collisions. ASPIS is easy to implement without requiring hardware changes while at the same time maintains security and can be incrementally deployed alongside legacy devices/infrastructure. We showcase the practicality of ASPIS by implementing it on the OpenAirInterface platform. Further, we demonstrate its effectiveness through extensive evaluations via a combination of small-scale experimental evaluation and large-scale, realistic simulations.