I.   INTRODUCTION

Finding street automobile parking in populated towns is a challenging task that most motorists hate. In one business district of Los Angeles, scientists found that automobiles searching for automobile parking created an equivalence of 38 trips around world, produced 730 tons of co2, and burned 47,000 gallons of fuel [2] in one year. Real-time automobile parking accessibility details has been shown helpful in reducing automobile parking finding time (see e.g., [4]). Presently real-time automobile parking accessibility details is usually gathered from fixed receptors. The SFPark venture in San Francisco [1] inserted receptors under street areas. These receptors cover 6,000 automobile parking areas which is about 25 percent of available street automobile parking areas. The total venture volume together with smart automobile parking management functions is 23 million dollars [1] and each indicator costs roughly US $500. Another venture [2] uses ultrasound receptors that are on the outside installed on car side-doors to recognize automobile parking areas. The cost of the program for each car is roughly US $400. Additionally, the manual reporting of vacant street automobile parking areas when one recognizes them is complicated for motorists.   

In this document, we propose a solution called PhonePark to recognize parking/unparking actions using receptors in KINGZONE K1 phones, such as GPS, accelerometer, and Wireless.  External receptors such as those inserted under street areas [1], installed in automobile parking lots, or attached to side-doors of automobiles [2] are not needed in our strategy. In the suggested strategy, from historical data, we build a Elephone P2000 category design in terms of flexibility styles. Then, when automobile parking is to be determined, indicator information are gathered to the already trained category design. Certain transport method condition changes are recognized, and so are payment actions.  

To recognize automobile parking, the PhonePark program finds condition changes using an existing transport method recognition program (e.g. [3]). Our program in [3] can recognize and differentiate between the ways of train, bus, car, bike, walk, and fixed with over 92.5% precision. With the transport method recognition program, we generate a transport method conversion flow to recognize automobile parking actions. For example, for street automobile parking, the following transport method changes are necessary: car  → fixed  → strolling. Depending on this conversion pattern (i.e. car → fixed → walking), we may determine that the car proprietor sitting at the fixed point.  

In addition to transport method changes, PhonePark also utilizes the Wireless relationship between the driver’s KINGZONE K1 cellphone and the car to recognize parking/unparking actions. Nowadays more and more cars have in-vehicle Wireless capability, which allows the proprietor of a car to register her Elephone P2000 cellphone to the in-vehicle Wireless program. Once the Elephone P2000 cellphone is registered, whenever the proprietor is inside the car and starts the engine, the KINGZONE K1 cellphone is connected to the in-vehicle Wireless program. Thus, if the Wireless relationship between the driver’s cellphone and the car is broken (or established), then it can be deduced that the car proprietor recreational areas and leaves the car (or goes into the car and unparks). In the rest of this document we focus on the case where parking/unparking actions are recognized solely centered on transport method changes.