How is the contextual in-car user experience evolving?

Vehicle HMI and infotainment are entering a phase where AI, contextualisation, and software‑defined architectures are reshaping how drivers and passengers experience the car. Rather than treating the cockpit as a static set of menus and controls, the next generation of systems are responding to who is in the vehicle, what they are doing, and the broader environment in which they are driving. This mirrors the evolution of the web in the early 2010s into personalised, context‑aware services that adapt to each user and device. 

In-car contextualisation has been on the horizon for many years, but historically many use cases were difficult to deliver without advances in AI. Combined with increased processing power and a more mature connected ecosystem, it is now becoming viable. As a result, new use cases are emerging that enable more personalised, proactive, and predictive vehicle behaviour, changing both day‑to‑day usability and longer‑term expectations of what a car can do.

What is contextualisation? In-car contextualisation refers to systems that adapt outputs or actions in real time, based on a high degree of situational awareness. This includes inputs such as user profiles, vehicle occupancy and seating positions, occupant activity, physical state, vehicle location and speed, environmental conditions, and the surrounding context (e.g. traffic, infrastructure, noise levels). Increasingly, this is being combined with a “scene”-based approach popularised by Chinese OEMs, where existing hardware and features are orchestrated into coordinated responses, enabling new and more holistic outcomes without requiring entirely new systems. 

Use of ADAS contextualisation in BMW iX3 (SBD’s 635 HMI Benchmarking report) 

To be effective, contextualisation must default to delivering outputs that are useful, relevant, and timely, without crossing into perceived intrusiveness. Striking this balance is critical. Systems must be configurable and continuously learn from user behaviour, allowing both user-led and system-led tuning to ensure outputs remain supportive, seamless, and largely invisible.

By combining multiple live data sources with persistent user profiles, contextual systems can reduce driver distraction by minimising manual inputs, lowering cognitive load, and reducing eyes-off-road time. They also enable automation of routine tasks and introduce moments of delight that can strengthen brand differentiation and user loyalty.

Positive and negative effects

Positive effects	Possible negative effects
More intelligent, responsive systems	Perceived creepiness or invasiveness if poorly calibrated
Fewer steps to achieve desired outcomes	Poor implementation can degrade the experience more than having no feature at all
Reduced manual input and cognitive load	Difficulty achieving consistent cross-ecosystem user profiles
Decreased driver distraction	Challenges handling unpredictable or chaotic real-world scenarios
More intuitive and adaptive interfaces	Data privacy and security risks
Increased personalisation	Risk of driver complacency or over-reliance
Potential for strong “wow” moments and brand differentiation	Dependence on connectivity and backend services

Implementation timeline

The following use cases illustrate how contextualisation is expected to be deployed across key in-car domains:

Domain	Today	Short‑term future	Long‑term future
HVAC, preconditioning, cabin mood, scenes	Preconditioning according to calendar, weather, and typical departure times (already in Tesla, BMW).	Using learned preferences to adjust HVAC, removing the need for manual interaction in typical everyday driving.	Cabin pre‑configuration using live data (e.g. external cameras) to judge body size and shape, minimising manual seat/mirror setup.
	Basic preference‑based HVAC presets (e.g. temperature, seat heating) applied on entry, with manual override.	Cabin pre‑configuration (seat/mirror, steering weight, HUD layout, media) according to profile.	Deeply coordinated cabin mood and richer “scenes” that adapt proactively to biometrics and driver state (stress, fatigue) while coordinating HVAC, seating, light, media.
	User selectable “scenes” available.	Cabin mood and “scenes” (lighting, fragrance, massage, sound, display themes) triggered by context such as time of day, location, route type, occupants, biometrics.	Cross‑ecosystem arrival/departure scenes (home-car) where cabin ambience and home ambience co‑adjust based on journey context and user routines.
Core controls and ADAS / infotainment core	Partial automation of some core controls: auto wipers, auto lights, auto start/stop, electronic parking brake.	Full automation of core controls including wipers, lights, start/stop, parking brake, drive‑mode selection, locking, mirrors, with manual controls retained.	Partial or full removal of some routine core control inputs in everyday operation under higher automation levels and supportive regulation.
	Basic ADAS personalisation via sensitivity settings and modes; manual toggling to avoid nuisance alerts.	ADAS personalisation and reduced intrusiveness via better driver‑state modelling and context‑adaptive warnings.	Mode‑aware, multimodal ADAS interaction, where the system explains and negotiates behaviour with occupants based on context and preferences.
	Static infotainment layouts with favourite shortcuts, but limited true adaptation.	Dynamic adjustment of core infotainment display that prioritises most used/relevant controls to “zero level”, demoting less‑used items.	Fully context‑adaptive infotainment surfaces that change structure according to driving state, occupants, role (private/fleet) and task.
Media, logins, device linking, digital life	Account‑based media (e.g. Spotify) with stored logins; authentication via passwords or phone pairing.	Auto‑curated or generative media based on route, traffic, mood and user profile.	Deep digital‑life continuity where calls, documents and media sessions move seamlessly between home, personal devices and vehicle.
	Phone projection (CarPlay/Android Auto) providing continuity of apps and some digital‑life integration.	Automated logins (Spotify, accounts) via biometrics and secure hardware tokens.	Multi‑device, multi‑user orchestration, with fine‑grained control over which accounts and data are active per occupant and scenario.
	Basic digital life integration in some ecosystems or isolated apps such as calendar.	Federated identity and token‑based access to digital life (calendars, messaging, productivity tools), with strict privacy controls and on‑device processing where possible.
Navigation, cluster/HUD, parking, charging, payments	Turn‑by‑turn navigation with live traffic; basic cluster/HUD overlays; simple rerouting.	Contextual navigation-ADAS integration, AR HUD, reroute explanations to the level desired by driver, leave‑time notifications, routing optimised to driver preferences.	Proactive trip orchestration: vehicle proposes departure times, route strategies, charging and rest stops and negotiates via VPA across devices.
	App‑based parking and payment; manual selection of parking/charging options.	Cluster and HUD outputs adapting to scenario (manual vs automated driving, complex junction, poor visibility).	Cross‑journey service orchestration: parking, charging, food and other services bundled into coherent “trip plans” per user profile and role.
	Plug & Charge in limited ecosystems; manual stop planning.	Automatic parking find/pay (on‑street, lots, hubs) based on preferences (cheapest, most convenient, charging available). Charging with Plug & Charge and wireless/robotic solutions, suggesting optimal stops including driver preferences, fatigue, calendar and route.	Charging fully embedded into car-home-grid optimisation, with explainable routing of charge/discharge decisions in the HMI.
	Location‑based scenes (e.g. lowering driver’s window automatically at specific locations, such as barriers).	Higher number of location-based scenes with greater complexity. Automatic payments for e.g. drive‑throughs via an in‑car wallet.
Biometrics, driver/passenger state, privacy	Camera‑based driver monitoring for drowsiness/distraction; fixed alert strategies.	Driver state monitoring for drowsiness and distraction including mood/stress classifications to adapt HMI complexity and cabin mood.	Continuous, multi‑modal behavioural and biometric sensing driving a holistic state model used across ADAS, cabin, media and notification policy.
	Static notification behaviour; manual configuration of what pops up when.	Adaptive notification strategies (suppressing non‑urgent messages under high workload).	Context‑aware privacy policies that change automatically with occupants and situation (e.g. stricter privacy with guests, “incognito” drives).
	Simple privacy settings per profile.	Escalation protocols (contact relatives/emergency services under certain biometric triggers, subject to regulations).	Transparent “why” surfaces that explain each contextual action and allow tuning or disabling entire classes of behaviour, including health‑related escalations.
	Basic audio zoning (front vs rear) in some vehicles.	Privacy/sound zones automatically tuned to use case (e.g. call vs entertainment vs children in back seats).
User profiles, passengers, VPA, lifecycle/shared use	Basic driver profiles (seat, mirrors, radio presets) per key/fob or account, with manual switching.	Biometric‑based user profile switching (face/fingerprint/voiceprint, phone proximity).	Adaptive shared‑use behaviour for car‑share, rentals and subscriptions: temporary profiles, selectable consent levels, automatic data wipe.
	Basic voice assistant integration for navigation/media/phone/HVAC/some body control.	Passenger‑specific zones (display and media per seat, child vs adult UIs, business vs leisure mode) linked to occupancy detection and known profiles. Generative AI that can hold intelligent conversations with a modest level of user understanding.	Rich, context‑aware multi‑user orchestration where occupants can negotiate vehicle behaviour (comfort vs efficiency vs time) via natural language. The assistant becomes multiple companion agents that know each user’s life story, evolve over years and anticipate needs before the user articulates them.
	Simple occupancy differentiation (seatbelts, child seats, basic driver recognition).	VPA multi‑modal interaction that changes verbosity and modality based on context (core preferences, driving vs parked, solo vs family trip). Complex interior sensing to detect multiple occupant profiles.	VPA acting as an orchestrator across vehicle, home and devices, maintaining user‑specific behavioural policies and preferences across contexts.
	Fleet/shared‑use handled mainly by policy outside the vehicle.	Role‑based UI modes (private vs fleet, ride‑hail, delivery) with different access rights, data retention policies and HMI complexity.

Summary

Contextualisation is fast becoming a defining layer of the in-car experience, but its success will depend on how carefully it is shaped around real user behaviour, expectations, and limits. 

Its opportunities come from making the right decisions at the right moments, quietly, consistently, and almost invisibly. Achieving this requires a detailed understanding of when context adds value, when it introduces friction, and how those boundaries shift across different users, journeys, and markets.

SBD's top five tips to enhance contextualisation

1. Maintain a unified context engine  The system should keep a structured view of who is in the car, what mode it’s in (manual, assisted, parked), where and when the journey is happening, and current driver workload. All contextual behaviour should rely on this view, rather than isolated “clever” features.

2. Prioritise cognitive load reduction over feature delivery Context should first decide what to hide, delay or simplify, such as non‑urgent prompts, secondary UI elements, and optional content. Extra convenience or ambience should only be added when workload is low. This keeps contextualisation aligned with safety rather than novelty.

3. Provide deterministic privacy controls and transparent explanations Provide a small number of clear controls such as “private drive”, “no health‑based actions”, or “no commercial offers”, and brief explanations of why the car acted in a particular way. Drivers should be able to understand and adjust contextual behaviour without searching through complex menu structures.

4. Treat occupant identity and role as core architectural inputs Driver/passenger identity and role (owner, fleet driver, guest, family trip) must directly influence layouts, defaults, data retention and permissions. The same contextual features then behave differently in a private car versus a shared or rental vehicle, without needing separate designs.

5. Launch scoped contextual loops and measure implicit feedback Start with a small set of clearly defined behaviours, for example learned HVAC presets, adaptive notification handling, and smarter parking/charging suggestions. Measure how often they trigger, how often users override them, and in what conditions, then refine thresholds and expand from there.

"As AI evolves HMI into adaptive interfaces and rich conversational transactions, real-time situational awareness becomes essential. OEMs that fuse AI, software-defined architectures and deep contextual understanding into intuitive, trustworthy experiences will define the next generation of automotive UX." Adam Jefferson - Senior UX Expert at SBD Automotive

How SBD can help

SBD Automotive can help benchmark your position against the wider industry and identify where action is needed most. To explore how these trends impact your strategy, architecture and supplier roadmap, get in touch with SBD Automotive for a deeper discussion. Email info@sbdautomotive.com