Proactive AI in private mobility
How smart systems can enhance the travel experience
AI technology is evolving very fast changing the way how we interact with machines.
Businesses across all industries are in a race to add smart assistants to their products and mobility companies don’t want to miss out.
- But is integrating a chatbot into a vehicle system really what users need?
- How should we imagine AI interactions when we’re away from our computer screens, driving a car, or riding a bike?
- How can we leverage the power of AI systems beyond just reacting to prompts and instructions?
Leveraging the power of proactive AI systems
Most current AI solutions resemble an input-output interaction, where the system analyzes and reacts to a set of human prompts.
The context is provided by the user, and the way we ask the questions affects the quality of the answer.
It feels we’re using a new technology with old mental models.
If we collaborate with a human companion or consultant, we would expect them to anticipate our needs and provide support without waiting for us to come up with the perfect prompts for our requests.
This is especially important in the mobility context, where:
- The conditions around the user are constantly changing
- The user is focusing on a primary activity, such as driving or riding, making every other activity secondary
So, how can we design proactive AI systems that provide real assistance to mobility users?
1. Start from real users needs
Like any technology, AI solutions can provide real value only if they address real user needs.
Imagine you’re a driver or rider.
What are your primary needs? What would you expect from a smart assistant?
You probably don’t start every journey talking about dinosaurs.
You’d rather be more interested in knowing how to get safely to your destination, avoiding traffic and bad weather conditions on the route.
You’d also expect a smart assistant to know your primary needs and anticipate them by offering timely suggestions.
But needs can change depending on the context. The better the system can analyze the context around the user, the more accurately it will be able to assist them.
2. Define the context
During a journey the context can change constantly.
Users might need to take action if different conditions or unexpected events arise. Internal and external factors can deeply affect the experience.
The context can be influenced by different factors:
User
- Which users are taking part to the journey?
- What does the system know about the user’s ecosystem (calendar, connected services, …)?
- What are the user preferences and habits?
Environment
- What are traffic and weather conditions around the user’s location? How are they changing along the route?
- Which services are available along the route?
Vehicle
- What is the battery state of charge?
- What are the vehicle settings? Are there any alerts?
Journey phase
- Is the user at the beginning, in the middle or at the end of their journey?
Context awareness is key to enabling systems to anticipate users’ needs and provide proactive assistance.
However, it’s also crucial to define when systems should take the lead and approach users, without overwhelming them.
3. Prioritize safety and relevance
While on a journey, focusing on what happens on the road is the primary task. Everything else is secondary.
Smart systems should assist the user while minimising distractions.
Safety first
- Keep information current and glanceable
- Keep interactions simple requiring low-cognitive processing
- Prioritise voice interactions where possible
- Human is in charge, AI should not be able to take actions by its own
Relevant information
- Safety information should always take the priority
- Smart systems should improve accuracy learning from human feedback
- Human is in control and can disable system’s suggestions at any time
- When AI’s suggestions are inaccurate, natural language interactions can help correcting the initial input.
4. Ensure data control and transparency
Proactive systems need to constantly run and analyse data on the background to suggest actions and predict outcomes for dynamic contexts.
This poses serious challenges around data privacy and control.
To ensure trust and transparency on data collection, proactive systems should be designed with:
- Clear feedback about the data being tracked
- Clarity around the benefits of enabling features connected to data tracking
- Easy access to view and manage tracking data
Proactive AI in private mobility: use cases
Let’s look at some examples of how proactive systems can anticipate and address user needs in different phases of a journey.
Before the journey
Before the journey the user wants to make sure everything is correctly set for the departure.
The system analyzes the context and inform the user if any action is required.
The interaction happens outside the vehicle on other devices, e.g. on the user’s mobile phone.
Use case example:
User need:
- I want to have enough battery charge before departure
Contextual data:
- Time of departure
- State of charge / Effective range
- Weather forecast
- Planned route
Scenario:
- Weather forecast with low temperatures ahead of a trip
Interaction:
- “Expecting -10° tomorrow. Set charge limit to 100% to increase effective range to 300km”
Starting the journey
At the time of departure the user want to be sure everything is set up for a comfortable trip. The system offers suggestions based on their habits.
The interaction now happens within the vehicle.
Use case example:
User need:
- I want to get navigation suggestions based on my habits / plans
Contextual data:
- Planned route
- Driving habits
- People in the car
Scenario:
- I’m usually going to gym on Mondays at 8:00 pm
Interaction:
- “Roadblock on the way to gym XYZ. View alternative route”
During the journey
During the journey interruptions and distractions should be reduced to the minimum. Other passengers can take over the interaction with the system while the driver is busy.
The system adjusts the suggestions following a natural conversation with the user.
Use case example:
User need:
- I want to be sure I reach my destination
Contextual data:
- State of Charge
- Location
- Route
- Nearby services
Scenario:
- Running out of battery before destination
Interaction:
- “4 fast charging spots available in 4 km. Cafe and playground nearby.”
End of journey
Once at destination users want to move away from the vehicle .
Further interactions should be transferred to mobile devices.
Use case example:
User need:
- I want to make sure my vehicle is charged before my next trip
Contextual data:
- Time of departure
- State of charge / Effective range
- Weather forecast
- Planned route
Scenario:
- Not enough battery to reach usual destination
Interaction:
- “You usually drive to work on Mondays. Schedule a charge overnight to have at least 50% battery before 8:00am”
TL;DR
AI technologies have the potential to transform the experience of private mobility, if they move beyond basic chatbots integration.
Proactive systems can anticipate user needs and offer personalised assistance at each step of their journey.
Key aspects when designing Proactive AI Systems include:
- Understanding primary user needs
- Defining and adapting to changing contexts
- Prioritising safety and information relevance
- Ensuring data control and transparency
I’m Pier 👋 I work as Design Lead at SnappMobile. I enjoy writing about Design and Mobility topics. If you enjoyed this post, feel free to leave a comment below or write me at @pierluigirufo. Looking forward to hear your feedback! 🙌
