Movement Detection using a Nearby Mobile Phone
Problem
There is a substantial amount of work in determining the location or activity of an individual over time inside a building with wireless tracking. When it is performed without devices, it is known as Device Free Passive Localisation (DfPL). This is where the human body causes a noticeable distortion to the wireless medium (unless the environment is very ‘noisy’). We have identified a unique (patent pending) mechanism to detect nearby movement through the use of mobile phones where software on the phone detects variations in the received signal power thereby allowing decisions to be made on whether a person(s) is moving in the vicinity of the mobile phone. This opens up the possibility of a multitude of applications making use of knowledge which can determine if a person is moving in the vicinity. This idea can also be easily transferred to other wireless devices such as laptops but the mobile phone is the more obvious candidate. At present the only way to determine human presence in a room equipped with a mobile phone would be to run some application which works through the camera on the phone. This would be a heavy drain on the phone battery as image processing is intensive. The camera on the phone also has to be orientated in a certain way and would only have a specific viewpoint. The other option might be sound recognition. Again, this is processor intensive and also prone to repeated failure should the person(s) speak no words.
Aim
This project will build on our UU patented technique which identifies movement without the need for the person to speak. Using mobile phones in a ‘device-free technique’ is different in that we are not attempting to ‘track’ an individual from a remote location but rather apply the principles of radio interference through the presence of a human so that a nearby phone can actually track the movement of a person for all manner of alerts and subsequent actions. Here the mobile phone is the ‘access point’ and the device which detects movement. This would allow app developers to incorporate this unique technique of detecting movement into their existing apps. One simplistic example could be alarm clock apps which detect the person being awake and moving in the room and thus cease to sound the alarm tone. Our technique is also useful in allowing a mobile to ‘be ready' as it detects movement e.g. screen of the mobile phone may turn on if movement is detected. The phone does not need to be docked and the threshold of detection can be adjusted so that disturbances by animals do not lead to false positives. The resolution of detection is room level ideally at less than 1 meter
NOTE: I have previous Java code for the system which can be used as your starting point.
References
[1] Gabriel Deak, Kevin Curran, Joan Condell and Daniel Deak (2012) Motion Detection using Device-free Passive Localisation (DfPL). ISSC 2012 - 23nd IET Irish Signals and Systems Conference, NUI Maynooth, Ireland, 28th-29th June 2012
[2] Gabriel Deak, Kevin Curran, Joan Condell, Nik Bessis and Eleana Asimakopoulou (2012) IoT (Internet of Things) and DfPL (Device-free Passive Localisation) in a disaster management scenario. Simulation Modelling Practice and Theory, Vol. 34, No. 3, pp: 86-96, DOI: 10.1016/j.simpat.2013.03.005, ISSN: 1569-190X, Elsevier Publishing