Global Position System (GPS) can be used to determine a geo-location of a GPS-enabled device. For example, a GPS-enabled device can receive GPS broadcasts from satellites orbiting 12,000 miles above the surface of the earth. There are challenges with the power requirements, reception of GPS signal, and accuracy of GPS-enabled devices. Today, GPS exists as a “single point of failure” and our dependence on GPS is increasing day by day. According to a recent study a GPS break down will cost American economy at least ONE BILLION dollars PER DAY. The proposed technology provides “Geo-Location” without satellite communication. We use sensor-actuated deep learning neural network to estimate geo-location. However, our Neural Networks require pre-training as well as on device training that requires a variety of supporting devices which can be grouped together as “NOGPS Technologies” which is an acronym of “Networked On-Ground Positioning Systems” Technologies. This can also be summarized as “AI + MEMS = NOGPS” (whereby AI = Artificial Intelligence, MEMS = Micro Electro Mechanical Systems).