Mansi Mehta — Product Designer

Remote Park Assist

Redesigning remote parking so drivers can move their car from outside it, with clear steps, safe controls and feedback they can trust.

Full casestudy

See it live

Remote Park Assist lets drivers move their Mercedes from outside the car using their phone — usually in tight garages or narrow parking spots.

I worked on the UX redesign of the driver interaction model, focusing on gesture control, live feedback, guidance, and blocked states.

Client

Mercedes-Benz R&D India

Team

Cross-functional (ADAS, HMI, engineering, QA)

Role

Product Designer · interaction model, UX research, prototyping

Platform

Mobile app controlling the vehicle (iOS + Android)

The Problem

The feature worked, but drivers still hesitated during use. The issue was not whether they could complete the task — it was whether the interface made each step clear enough to act on.

Gesture was the biggest barrier

Drivers had to press-and-hold while tilting the phone. Many weren’t sure whether to press first, tilt first, or how far to tilt.

Onboarding came too early

Instructions appeared right after authorization, before drivers had enough context. Many skipped it and then struggled during the actual gesture.

Feedback didn’t confirm enough

Drivers couldn’t always tell whether the app had registered their input, whether the car was ready, or whether an environment condition was blocking progress.

The through-line was clear

People could complete the task, but the interface did not consistently show what was happening, what had changed, or what action was expected next.

The most interesting parts live in the full walkthrough.

Request walkthrough

View on store

Core Decisions

Few decisions shaped the redesign: make gesture feedback readable at a glance, treat interruptions as part of the main flow, and move guidance closer to the moment of action.

Gesture

Instead of adding more instruction, I designed a progressive fill that grew as the gesture was performed and completed at the threshold.

The driver could keep looking at the car and still catch, through peripheral vision, when the system registered the gesture and cleared the car to move.

The fix was not more text. It was feedback that confirmed progress without demanding full visual attention.

Gesture feedback inspired by familiar motion controls

Guidance

The earlier flow explained too much before drivers had real context. People skipped it, then hesitated during the actual gesture.

I moved guidance closer to the moments where hesitation happened. Lightweight prompts supported the core control, while Top View helped drivers understand the car’s surroundings without leaving the flow.

Error system

The ideal flow was simple: connect, authorize, move, complete. But the critical moments were the interruptions.

I mapped and designed over twenty blocked and recovery states. Each state explained what happened, showed whether recovery was possible, and gave the next action.

Process

I studied first-time and experienced drivers across India and Germany to understand where the interaction broke down. The research focused on:

how drivers held and moved the phone
where they hesitated before motion started
which feedback they noticed or missed
how they recovered from blocked states

The same patterns kept appearing: gesture uncertainty, skipped onboarding, unclear feedback, and weak recovery moments.

Those patterns became the foundation for the redesign decisions.

Challenges & Learnings

The hardest part was that I couldn’t remove the core gesture. The deadman control was a safety requirement, so the design challenge was to make the existing mechanism clearer, more predictable, and easier to recover from.

The second challenge was feedback. Early versions explained too much. The better direction was to confirm only what mattered: input registered, car ready, blocked state, next action.

The biggest learning was that safety-critical UX is not just about reducing steps. It is about reducing uncertainty at every point where the user has to act.

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

iOS + Android

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

iOS + Android

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

iOS + Android

See Other Projects

Remote Park Assist

Redesigning remote parking so drivers can move their car from outside it, with clear steps, safe controls and feedback they can trust.

Full casestudy

See it live

Remote Park Assist lets drivers move their Mercedes from outside the car using their phone — usually in tight garages or narrow parking spots.

I worked on the UX redesign of the driver interaction model, focusing on gesture control, live feedback, guidance, and blocked states.

Client

Mercedes-Benz R&D India

Team

Cross-functional (ADAS, HMI, engineering, QA)

Role

Product Designer · interaction model, UX research, prototyping

Platform

Mobile app controlling the vehicle (iOS + Android)

The Problem

The feature worked, but drivers still hesitated during use. The issue was not whether they could complete the task — it was whether the interface made each step clear enough to act on.

Gesture was the biggest barrier

Drivers had to press-and-hold while tilting the phone. Many weren’t sure whether to press first, tilt first, or how far to tilt.

Drivers had to press-and-hold while tilting the phone. Many weren’t sure whether to press first, tilt first, or how far to tilt.

Onboarding came too early

Instructions appeared right after authorization, before drivers had enough context. Many skipped it and then struggled during the actual gesture.

Instructions appeared right after authorization, before drivers had enough context. Many skipped it and then struggled during the actual gesture.

Feedback didn’t confirm enough

Drivers couldn’t always tell whether the app had registered their input, whether the car was ready, or whether an environment condition was blocking progress.

Drivers couldn’t always tell whether the app had registered their input, whether the car was ready, or whether an environment condition was blocking progress.

The through-line was clear

People could complete the task, but the interface did not consistently show what was happening, what had changed, or what action was expected next.

People could complete the task, but the interface did not consistently show what was happening, what had changed, or what action was expected next.

The most interesting parts live in the full walkthrough.

Request walkthrough

View on store

Core Decisions

Few decisions shaped the redesign: make gesture feedback readable at a glance, treat interruptions as part of the main flow, and move guidance closer to the moment of action.

Gesture

Instead of adding more instruction, I designed a progressive fill that grew as the gesture was performed and completed at the threshold.

The driver could keep looking at the car and still catch, through peripheral vision, when the system registered the gesture and cleared the car to move.

The fix was not more text. It was feedback that confirmed progress without demanding full visual attention.

Gesture feedback inspired by familiar motion controls

Guidance

The earlier flow explained too much before drivers had real context. People skipped it, then hesitated during the actual gesture.

I moved guidance closer to the moments where hesitation happened. Lightweight prompts supported the core control, while Top View helped drivers understand the car’s surroundings without leaving the flow.

Error system

The ideal flow was simple: connect, authorize, move, complete. But the critical moments were the interruptions.

I mapped and designed over twenty blocked and recovery states. Each state explained what happened, showed whether recovery was possible, and gave the next action.

Process

I studied first-time and experienced drivers across India and Germany to understand where the interaction broke down. The research focused on:

how drivers held and moved the phone

where they hesitated before motion started

which feedback they noticed or missed

how they recovered from blocked states

The same patterns kept appearing: gesture uncertainty, skipped onboarding, unclear feedback, and weak recovery moments.

Those patterns became the foundation for the redesign decisions.

Challenges & Learnings

The hardest part was that I couldn’t remove the core gesture. The deadman control was a safety requirement, so the design challenge was to make the existing mechanism clearer, more predictable, and easier to recover from.

The second challenge was feedback. Early versions explained too much. The better direction was to confirm only what mattered: input registered, car ready, blocked state, next action.

The biggest learning was that safety-critical UX is not just about reducing steps. It is about reducing uncertainty at every point where the user has to act.

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

OUTCOME

Live in all markets

The final release brought together the full RPA redesign from onboarding and gesture control to live feedback, Top View guidance, blocked states, recovery paths, and exits.

I owned the mobile control flow end to end. Real product screens are not shown due to NDA.

Curious about the full flow or want to dive deeper?

Request walkthrough

See Other Projects