Connected Carcierge

Inspiration

Buy with us, service with us, trade with us, repeat

Visiting a service department needn't be an ordeal. Connected Carcierge proposes to make it an experience worth repeating, with timely service notifications, accurate scheduling and estimation, helpful customer education, appraisal and trade integration, and a personalized in-dealer experience befitting a valued customer.

What it does

Not just a service kiosk, but an everywhere-at-once service assistant

Carcierge Mobile

In-app and in-car service appointment scheduling, via a friendly voice-controlled dialog, allow customers to see to service needs as they arise. Suggestions of appointment time and location as well as services to be performed take into account the customer's schedule, favorite dealership, and budget. Additional offers such as service specials, recommended warranty repairs, and parts and accessories can be made as the appointment is scheduled. Real-time updates from the service desk let the customer know how things are going, and when it comes time for a new set of wheels, an accurate trade-in offer is just a voice query away.

Carcierge Connected Desk

Thanks to geo-fence technology, the service desk is alerted when a customer is close to the dealership. If Internet connected sensors find that the service bay is full, all the service manager needs do is assign the to a different one, and the customer gets a voice message letting him or her know where to pull in. Once the customer arrives on site, the service manager is alerted to the assigned service bay and personalized kiosk. If service personnel need to speak to the customer directly, they just meet at the assigned kiosk, where the customer is likely already sipping on his or her favorite beverage while learning more about the service being performed.

Carcierge Kiosk

More a nook than a kiosk, really, we envision this as being a downright pleasant place to be. Upon arrival, the customer checks into the assigned kiosk, and is greeted by name thanks to either a sensing beacon or facial recognition, offered a favorite refreshment, and invited to settle into a comfortable reclining chair for the duration of the service appointment. In the mood lit kiosk is an easily accessible, voice and touch enabled panel where the customer can learn more about the service being performed, browse parts and accessories, be apprised of service offers, and offered a free trade-in inspection.

If the customer just wants to relax and listen to some music or catch a show, that's fine too, but we feel many will leave with a trade-in estimate in hand. And once the estimate is in their profile, they can ask for and receive it every time they cruise by a dealership and see a hot new sports car on the lot. And why not come back for all service from that point forward, when the best trade-in offer is guaranteed thanks to an "in house", comprehensive service history?

The Data

One feature we didn't get to cover too much during our demo is the massive reporting and analysis opportunities afforded thanks to all the data we'll gather. Beyond just coffee preferences and service frequency, the customer profile can feature typical driving habits, routes, commute times, and the like. Uses for just these points range from service and accessories recommendations to advising on the purchase of a new car. A mountain commuter might like better shocks, or a vehicle with all wheel drive, say, while more spirited driver might be happiest with a sports car or some racing tires.

In addition, there's dealer performance data to consider. How good is a particular service department about getting its customers in at scheduled intervals? How many bay conflicts are there due to service jobs running over schedule? How much educational material are customers consuming at the kiosks? Do we have enough decaf on hand?

How we built it

For the hackathon, we implemented an in-vehicle appointment scheduling dialog triggered by Node-RED connection set up via Bluemix. The connection served is a stand in for any number of current of future Internet-connected trouble reporting systems. Once Carcierge was alerted to the trouble, it began a scheduling dialog with the driver. The client itself was a slight modification of the very handy Watson conversation service demo, a Node.js application which (conveniently!) depicted a connected car. Behind the scenes, Watson's conversation, speech-to-text, and text-to-speech services were employed to facilitate conversation.

Challenges we ran into

We'll need to write a "Teamcierge" before the next hackathon that entices team members to show up. Only two of our originally pledged six members showed up, and only one of those was able to commit to a full-time presence. We were able to pick up a very motivated "intern", however, whose enthusiasm helped keep things moving. In addition, Bluemix is a very capable platform with which we should have spent more time before the event. Special thanks must go to Ram, the ambassador from IBM, for helping us through the ins and outs of the platform and especially for the Node-RED piece.

Accomplishments that we're proud of

The idea itself has merit, we feel. Our demo wasn't stellar thanks to the challenges we faced, but given time to refine and implement, we Carcierge it has legs. The talking-car demo was cool, if not completely novel.

What we learned

Even when working from fairly detailed requirements, there's a lot of room for creativity. We weren't sure where to go with the kiosk idea at first, but then we incorporated it into a broader customer experience and it fit very well.

What's next for Carcierge

We'd like to develop the product a further. There's more work to be done on a prototype of Carcierge Mobile, to be sure, and a lot of opportunity for cool development on Connected Desk and Kiosk. Of the three parts, Kiosk has the potential to be the "crown jewel", delivering an experience that may have customers referring their friends to a dealer for an oil change just to try it out.

Built With

• bluemix
• data
• geo-fencing
• ibm-watson
• iot
• node-red
• node.js