STELLANTIS 2030 REDESIGN
-
Industry Benchmarking (ScreensStudio)
-
Vehicle Sketching
-
Driver Cluster & Center Console
-
Multimodal Experiences
-
Driver Accessibility and Tendencies
How might we use AI & Machine Learning to reduce range anxiety and improve overall vehicle performance by managing energy smartly and predictively?
Overview
As part of my Automotive UX class at the University of Michigan, SI 311.150, we partnered with Stellantis to imagine what Automotive User Experience could look like in vehicles of the future. My partner and I were most interested in their second prompt: "How might we use AI and Machine Learning to reduce range anxiety and improve overall vehicle performance by managing energy smartly and predictively?". As EV infrastructure becomes more of a priority in today's vehicle landscape, we ideated features for the Jeep Avenger 2024 that would streamline the EV charging process and predictively and proactively address needs of novice EV drivers in the near future. Some of our proposed implementations included integrating calendars into main vehicle functionalities, using AI alerts and suggestions to assist battery management, and optimizing a driver’s schedule even before they enter the vehicle.
The Problem
The primary concern for many novice EV drivers is the hassle of scheduling their day around charging or not having sufficient range to reach their destination.
Objective
After a thorough exploration of the problem space and benchmarking of current industry successes, my partner and I sought to incorporate AI as a means of automating charge schedules, reducing range anxiety, and teaching healthy charging habits to optimize battery health.
Timeline
-
10 weeks (Oct 2024 - Dec 2024)
Tools
-
Figma - Prototyping, Tracing, Presentation
-
ScreensStudio - Vehicle benchmarking, HMI research and standards
Demographic Research
Charging Woes
During our preliminary demographic research, we honed in on a specific type of EV user: novice and first-time electric vehicle owners. After conducting some short surveys and reviewing external sources, we quickly identified that the majority of this demographic do not have confidence in the capabilities of the current EV infrastructure. This, in turn, has led to a measly 18% of people reportedly willing to buy EVs in 2024, while 50% of current EV owners exclusively charge their vehicles at home. These numbers reflect an underlying anxiety related to the hassle of charging on the go. With the hustle and bustle of everyday life, drivers don't want to add another inconvenience to their already busy schedules. Through our redesign, my partner and I hoped to address some of this reluctance to drive EVs.
People fear the prospect of getting stuck with no charge while en route to their destination.
Design Overview
Interior sketch of Jeep Avenger with Driver Cluster and Center Console Designs
Style Guide
Key Design Decision #1
Calendar Integration
One of the most important ways we chose to incorporate AI effectively into the EV energy management experience was by automating charge scheduling so that it populates in a driver's calendar without them having to even think about it. Most users already have their mobile devices and calendars connected to their car. By analyzing and learning the patterns of the driver's typical daily and weekly schedule over time, AI-suggested charging events can be inserted into the most logical slot in the calendar, taking into consideration battery level, charging location, and free time. As a driver's schedule inevitably updates and changes, the live-edit suggestions can adapt and move with the new schedule updates. This creates one less step for the driver to charge their vehicle while on the go, ensuring their charging will not derail their plans, but rather, optimize their time efficiency and battery health.
.png)
Center console landing page with expanded calendar widget. Suggested events populate in the calendar and can be accepted or declined.
Multimodal Scheduling
The second aspect of calendar automation that we focused on was creating a multimodal scheduling experience. Rather than the user experience starting as the driver enters the vehicle, we believe range anxiety can be further reduced by creating an experience that starts from the comfort of one's own home. By using data collected from the driver's smart home devices, such as grocery lists, scheduled events, upcoming deliveries and more, the vehicle's accompanying mobile app can predictively plan when the vehicle is in need of a charge. It can then schedule the charge to align with other upcoming events, such as grocery shopping, saving the driver from the hassle of coordinating errands and efficiently organizing calendar events.
Multimodal scheduling scenario between Google Home and accompanying app
Key Design Decision #2
AI Alerts
In addition to automating scheduling, we felt that it was important to teach novice EV drivers about good charging habits, as well as best practices for preserving battery health. To achieve this, we implemented an alert system in the vehicle that serves as 'training wheels' to teach these habits actively and instantly. Drivers have the option of disabling or enabling these alerts depending on their level of comfort and experience in the vehicle. These alerts are available in a multitude of places, including the driver cluster, center console, and mobile app, in order to reinforce the message for the driver. While they may be closed out immediately from the steering wheel or center console button, the alerts will disappear on their own after a brief window of time to minimize distraction.
Driver cluster with alert popup notifying driver that they are braking too abruptly. More information can be found in the accompanying app as seen below
AI Suggestions
While we have already ideated for users who are very organized and schedule-oriented, we recognize that many drivers, especially within the Jeep family, are more spontaneous in nature, and may not even plan out where they are going before they start driving. For these driver personas, we implemented an AI-powered "Add Charging Stop" feature, which can populate several options for charging stops along a given route, even as the user is driving. This ensures the least amount of disruption and distraction so that their eyes never leave the road for more than a few seconds.
Center cluster navigation page with suggested charging stops along a planned route.
Key Design Decision #3
Pre-Drive Optimization
Another way to eliminate anxiety and prevent the driver from feeling overwhelmed is to utilize AI learning to optimize driver comfort before they even enter the vehicle. We have implemented several controls that can adapt to the driver's ideal settings, from the draggable climate app, vehicle controls, and mobile app control interface. All of the settings can be adjusted even when the driver is away from the vehicle, and can adapt to battery level, weather conditions, driver mood, and time of day. Drivers can also remote start their vehicle and adjust profile preferences from the mobile app.
Voice Control
Implementing an AI voice assistant for the vehicle will allow drivers to be notified about all of the suggestions and decisions being made for them without having to take their eyes off the road or pressing any buttons. This can also relieve their mental load by knowing that an AI assistant is present to deal with charging logistics and answer any questions.
Voice control enables hands-free interaction while confirming the AI decision-making in the vehicle
Future & Reflections
Next Steps
Given that this project was completed within a semester, there are several steps I would consider taking to further the potential of AI incorporation, especially considering the external features of the car that can make driving a more holistic and multimodal experience.
