Foot Gesture Interface for LoRa Emergency Systems

Sammanfattning: Abstract A prototype for a smart shoe has been developed at the Royal Institute ofTechnology. It consisted of an Internet of Things (IoT) device which was builtand integrated with a radio module to wirelessly transmit packets. With thissmart-shoe as a base, this thesis attempts to create a system that has the ability torecognize certain foot gestures, and utilizing them to trigger a radio transmissionwhen faced with a danger to hopefully receive the aid that is necessary. Thedangers could for instance be in the event of abuse, kidnapping or robbery. Inorder to accomplish this, machine learning (ML) models were used to predictwhat activity or gesture was executed. For the radio communication, LoRa(Long Range) was utilized as the modulation technique and enabled wirelesstransmissions. A system was developed that can detect two gestures with a high accuracy.Whenever a gesture is performed, the radio module is activated and generatesa transmission to another unit. Keywords Internet of Things, embedded systems, radio communication, LoRa, machinelearning, foot gestures, recognitioniv