Project Background

Melbourne Transport Community Project provides trains and tram departures to the end user, over new channels.

“The challenges people have when travelling are numerous”, said Fabio Douek, Head of Cloud Delivery, Singlepoint. “When commuters want to check the timetable, the only channels available to the general public are the traditional web, mobile application and call centre. These modes of accessing available information don’t suit everyone. We wanted to provide a voice activated service on digital personal assistants so that information was available which could be accessed by everyone”.

People who are home with their hands full, perhaps looking after children, or people with disabilities find voice activated personal assistants, such as Alexa and Google Home to be a good alternative. Visitors to Melbourne who may not want to download the mobile app find chatbots very user friendly. Most people have Facebook on their phone these days and other messenger applications can be used such as Viber, Telegram, Skype, Slack and Twitter. “The reason we support all those different messenger applications is because we believe that the end users know which one is best for them”, continues Fabio. “What we see is that, in some cases, a user uses different channels depending on where they are. For example, using an Alexa Device from home, Facebook messenger from their mobile phones when they are on the go, and Slack when they are just about to leave the office.”

case-study-melbourne-transport

Work Delivered

Singlepoint developed the microservices which run on the AWS platform. On top of that, three different solutions were used to provide the integration with the devices for the NLU (Natural Language Understanding): Amazon Alexa, Amazon Lex and API.AI. These support Alexa enabled devices, such as Amazon Echo and Echo Show, Google Home, Facebook, Slack, Skype and Telegram. Singlepoint also delivered a Proof of Concept (POC) for native Mobile Apps which allows the user to interact with Melbourne Transport either by text message or voice. Another POC was delivered, integrating Amazon Lex with Amazon Connect, which allows users to call a contact centre (by either using a mobile phone or landline) and interact using the phone keypad or voice.

Results

Melbourne Transport Community Project is a non-profit initiative led by Singlepoint. “We decided to kick-off this initiative for two main reasons,” adds Fabio. “The first one was to give end users more flexibility and freedom on how they want to consume public transport departure information. The second motivation was to create a platform where we could experiment with the emerging technologies and innovate rapidly. The main goal is to use our know how to deliver value to organisations who want to innovate. The Melbourne Transport community project has helped us in this journey. We are able to show the possibilities to prospective customers which they are then able to correlate to their own use cases.”

“Currently the chatbots are taking advantage of NLU provided by Alexa, Lex and API.AI which handle specific use cases e.g. Next train from to .” concludes Fabio. “This meets our expectations but the next phase will bring Machine Learning into the conversations. For example, I might want to ask: ‘What time shall I leave so I will be in the city at 08.30?’ If the user takes the train in the morning to go to work from Monday-Friday, Machine Learning could recommend that route, or even push notifications based on the channel.”

Start the conversation today!

See how our team can support your organisation