Using machine learning to detect pedestrian locomotion from sensor-based data Courtney Anne Ngo De La Salle University Manila, Philippines courtney_ngo@dlsu.ph Solomon See De La Salle University Manila, Philippines solomon.see@gmail.com Roberto Legaspi roberto.legaspi@gmail.com ABSTRACT The integration of low cost microelectromagnetic (MEM) sensors into smart phones have made inertial navigation systems (INS) possible for ubiquitous use. Many research studies developed algorithms to detect a user’s steps, and to calculate a user’s stride to know the position displacement of the user. Subsequent research have already integrated the phone’s heading to map out the user’s movement across a physical area. These research, however, have not taken into account negative pedestrian locomotion, wherein the user is moving but is not exhibiting any position displacement. Current INSs are not suited to handle negative pedestrian locomotion movements, and this leads them to consider false steps as real steps. As the INS’s modules depend heavily on the outputs of the other modules, a cascading error would most likely occur. This research aims to solve this problem by collecting pos- itive and negative pedestrian locomotion with data from phone-embedded sensors positioned in the research subject’s front pocket. Using these data, a model will be built to clas- sify negative pedestrian locomotion from positive ones, and to eventually improve the INS’s accuracy overall. Categories and Subject Descriptors H.4 [Location-based Services]: [Information Systems, In- formation Systems Applications, Spatial-temporal Systems] General Terms Machine Learning, Inertial Navigation Systems, Sensors 1. INTRODUCTION Indoor navigation systems determine where a device has tra- versed inside a building. These navigation systems can be employed in applications to help users find a specific location in closed places like conference centers and office buildings. Unlike outdoor navigation systems like the Global Position- ing System (GPS), indoor navigation systems can not use satellite signals as heavy attenuation takes place when the signals make their way through physical obstacles. To solve this, researchers have experimented with Wi-fi sig- nals [2, 1, 18, 19, 3], vision [6], ultra-wide bands [15], cellular- based signals [11], magnetometers [5], and combinations of these [4]. All of these research are dependent on environ- ment variables such as Wi-fi routers and markers, and some require data collection prior to system use. This would mean that a significant change in the environment or the variables would affect the performance of these navigation systems. INSs, on the other hand, uses data from inertial sensors such as gyroscopes and accelerometers to determine the path a device has travelled. Smart phones currently already have these sensors as micro-electrical-mechanical systems (MEMS) devices, making it possible for INSs to be applied in smart devices and possibly for ubiquitous use. Compared to other navigational systems, INSs are independent of its environ- ment, requiring less cost that otherwise would have incurred with the need of access points. This also implies less envi- ronment set-up as access points do not need to be installed for the navigation system to operate. Considering that it is a cheaper and simpler alternative, INS appears to be a more attractive approach to building navigation systems. Challenges Using INSs in real-world situations, however, is limited be- cause its MEMS devices are susceptible to noise and grad- ual drifts that cause cascading errors. Because of this, most existing INSs integrate regular checking with access points with known positions such as satellites and Wi-fi routers to calculate the position of the mobile unit to compensate for these inaccuracies [9]. Another problem, which this study intends to address, is correctly classifying irregular movements. In this research, positive pedestrian locomotion is defined as movements that include moving from one physical position to another on foot. Examples of these are walking, jogging, running, and climbing up and down the stairs. False pedestrian loco- motions are movements that do not require moving from a position, such as standing. There are, however, some false pedestrian locomotion movements that can simulate move- ment from position, and these presents a problem to some existing INSs. These movements include walking-in-place, jogging-in-place, and running-in-place. It is important fu- ture INSs can correctly disregard false pedestrian locomo- tion movements to avoid cascading errors as the modules depend on each other as displayed in Figure 1. Similarly, it cannot be expected that users would not exhibit any form of negative pedestrian locomotion movements in real-world applications. An INS that considers in these negative move- ments will better suit mobile applications that plan to map user paths in an area.