Machine Learning Supported Assisted Travel

Travel for work or education is a significant challenge for many people. There are one million young people aged 16 - 24 with anxiety, 3.5 million people with Sensory Processing Disorder and 700,000 who are autistic. Many of these people struggle to fully engage with society due to the difficulties they face on the journeys they must undertake to participate in work and education. BOM has created an app Hidden Kingdoms that uses spatialiszed audio and imaginary three dimensional worlds, where immersive Kingdoms 'mapped' onto travellers' journeys, support them as they experience anxiety or distress during commutes to work, school or college.

Testing of the app has revealed a limitation to the approach, and a significant opportunity. The initial overhead required to configure journeys in the app is an inhibitor to app use. Our project aims to reduce this overhead by exploring the possibility of using mobile phone, smartwatch or fitness tracker data to infer emotional state. Mobile devices are capable of capturing a wide range of data from sensors such as GPS, accelerometer or photoplethysmography (PPG) sensors that can be used to determine the traveller's gait (pattern of walking), direction of travel or heart rate. Machine learning approaches can be used to identify patterns in this data that indicate anxiety or distress.

Once identified the app can deliver relevant content, such as a friendly animal companion appearing and offering help, an invitation to view the Hidden Kingdom landscape on their phone screen, or therapeutic audio content being played back. The objective is to create an application that responds sensitively and appropriately to users' experience during their journeys, planned or otherwise.