Wrapping up our series of articles about our involvement with London Midland Labs, our solution architect Henry Todd describes our proof-of-concept, as well as our next steps.
Participating in the second London Midland Labs cohort has been a pleasurable challenge for us; we hadn’t previously been involved with rail travel and there’s a great deal of complexity which, all of us as travellers, are usually unaware of. Thankfully, we’ve had wonderful support from London Midland, and this hard work over the last 12 weeks culminated in a win at the Demo Day two weeks ago.
We demonstrated our vision for LoMi as a conversational interface to London Midland's services. Voice and text conversations are both equally capable of "flattening" interfaces — allowing a user perform a task very succinctly that might take multiple interactions in a graphical interface — and yet there are those tasks which are better suited to text, and a visual medium, and those which are best suited to voice.
Take booking a ticket as an example. It was clear to us (and borne out in testing) that a vocal conversation is significantly less effort than selecting your departure & destination stations from dropdowns on a website. Being able to say "I need to get to Birmingham", and being asked "when would you like to travel?", is a powerful flattening of the interface to ticket purchases and which consequently lowers the barriers to sale. Similarly, saying aloud a date & time for travel is far quicker and more natural ("tomorrow, at 2ish?") than selecting your travelling day from a calendar popup, a time from another dropdown, and so on.
However, as we soon discovered during development, there are often many trains running from any given station around busy commuter times. There are also many possible options for tickets (would you like a single or return, off-peak or any-time, first class, or perhaps you have a railcard?). Here is where we found it best to switch from a voice interface to a visual one. Knowing the details of your desired journey, LoMi can send a link to your phone via Messenger, allowing you to complete the booking process in a couple of taps.
There are many other possible interactions that work equally well via either voice or text. Asking if our train is on time, for instance, is something we found ourselves doing just as much via Messenger as via the Echo Dot in our kitchen. Here, it's worth noting a typical subtlety that can occur with conversational interfaces: context. Which train do we mean when we say "my train"? With LoMi we allow for both tickets you have pre-booked (either one-off or season ticket), and your most travelled route. There is much potential in those two words — "my train" — for genuinely useful applications of the data that London Midland has about its travellers.
Away from booking tickets and timetabling, there are many questions that travellers often ask: "can I take my bike?", "when does off-peak start?" and "can I use my Oyster card?". LoMi can answer these as part of an on-going conversation with customers, providing immediate responses via common channels like Facebook & Twitter.
Helping to shorten response times like this for London Midland staff, and provide a conversational view across all of these channels to assist with customer queries, is the next step. We're excited to be applying our expertise to a new domain, and we invite you to follow us as we do.
Thank you to everyone who came along to watch our demo & presentation, and especially to the staff of London Midland who have helped us learn so much in these 12 weeks.
About the Author
Henry Todd is part of the architecture team at Lola Tech, where he is responsible for translating project requirements into an effective solution design. Lola Tech is a pioneering travel software business leading the field in travel tech innovation. With big-name clients globally, their talented team is adept at creating awesome systems to increase your uptime and conversion rates or to provide you with that innovative edge to outshine the competition. Supporter of Dazzle, a voice activated personal assistant for the hotel industry. Visit Lola Tech for more information.