Concept 3 - Connection with other drivers on the road

by sharing the beats on the roads

The Concept

The cars will be able to display what the driver is currently listening on the exterior of the car. The beats of the song will be visualized on the outside of the car whenever the car stops due to traffic or a red light.

The drivers around will be able to listen to these songs with the help of their AI. When you stop the car, the AI of your car will suggest the songs that the other cars are listening according to your mood. Your mood will be interpreted by some face recognition systems. AI will show the suggestions on the front window and the driver will be able to “grasp” the songs if they like the beat visualizations. If they don’t like, they can pass on to a different suggestion (again made by the AI).

📋 Our Plan - What we wanted to achieve

In the third concept, we planned to use Arduino and Edge Impulse. Edge Impulsewill be used for collecting the input and Arduino & Voiceflow will show the output.

The input will be the movement (acceleration) data of the car (the seat that the participant sits) and it will be collected through the accelerometer on the phone. Through defining 2 different states: movement and stopping, the AI was able to differentiate between if a car was standing still in a traffic or accelerating on the road.

We planned to show the output when the car was at the stopping state. The output would be the colored beat patterns shown on the back of other cars, vibration on the driver seat and the AI asking whether they would like to listen the beats that they see or not. The colored patterns were planned to be visualized through 64 digits LED lights connected to Arduino. Vibration will also be actualized through a vibration motor connected to the Arduino. And the AI will be simulated through VoiceFlow.

Our plan is to connect Edge Impulse to Arduino so that when we receive stopping data, the Arduino and VoiceFlow will be activated.

📋 Our Plan - What we actually did

While prototyping, we have connected Edge Impulse to our phone so that it would collect the data of the phone’s accelerometer. Through this way, we could position the phone wherever we wanted on the physical prototype and get the relevant data. We trained the model by mimicking 2 different stages: stopping and moving.

Meanwhile, we also built Arduino board with the 64 digits LED lights and implement the LED library to mimic the beats pattern. We used 2 different patterns to mimic 2 different songs.

We have also created a simple AI speak flow in VoiceFlow in order to imitate the voiceover of AI during the selection process of the songs.

We couldn’t implement it in this prototype but in the future, we would also like to add the vibration sensors as one of the feedback from car to mimic the seat vibration.

In the end, we could not fully connect Edge Impulse with the Arduino but we used them separately. The phone, which was connected to Edge Impulse, was stating whether the car was moving or stopping. When it stated that it stopped, we used Wizard of Oz to manually enter the Arduino code to the led lights and let them display a colored pattern. We have also manually activated the Voice flow at this stage and directed it ourselves depending on the answers of the participant.

As a further step, we are planning to connect all of these separate parts together to make a fully working prototype.

🧙 Exploration by prototyping tests

Findings:

1. Participants felt connected to the other cars in the traffic

2. It was seen as a way to make the traffic more interesting

3. There could be moments when the driver wants to be left alone and not interact with any others

4. Seeing moving patterns and colors at the back of cars could be irritating and distracting after some point especially if you are not in the mood of interacting with other cars

5. Participants couldn’t really understand when did the colors appear or when did they turn off again (what triggered these)

Insights:

1. The interaction should be two-way in order to form a real connection with the drivers around. This could be done by signaling the other driver that the car behind wanted to listen their song/ liked their song choice/ wants more song recommendations from them.

2. If the driver says “no” to song requests many times, the the AI could ask whether they want to continue searching for music.

3. The driver should be able to turn on/off the AI (speaking to them) depending if they want to interact with the other cars or if they want to listen to their own music or maybe no music at all.

4. Also the ability to see the colors should be able to turn on-off so that it wouldn’t annoy the drivers if they don’t want to see the beats of other cars

5. The frequency/ pattern/colors of the lights should change depending on the beat of the music so that the drivers would have a better idea what could the music be like.

Why is it meaningful?

Currently, we face many instances where we lose our temper, or are confused with other drivers since we lose sight of them, as they are hidden inside a steel box. In the future, we feel that this phenomenon will increase, as AI will ‘improve’ driver’s behavior. Although this makes the roads safer and efficient, the driving behavior, and by extension our perspectives of each other on the road, will be less human.

At the same time, we feel that the car is a personal space to spend private time in, and we don’t want to feel exposed to strangers.

Listening to music is one of the most common ways to spend time in the car. Since it is audial, it does not keep the drivers’ eyes off the road. In addition, you can maintain your personal boundaries while showing personal preferences.

Sharing the beats on the roads would be a way to shift the drivers attention from the steel vehicles around to actually the humans inside. The interaction with the other drivers on the road will make them feel connected to each other. This way, they may escape from the chaotic atmosphere of the traffic and have a surprising and fun ride. It is also a way to increase the music knowledge of drivers and introduce new songs to them which they may like.