The Challenge

Traditional subsea operations rely on manual control of remotely operated vehicles (ROVs).

Pilots must simultaneously:

• monitor multiple sensor feeds

• manually control vehicle movement

• interpret complex subsea environments

As autonomy becomes possible, the operator’s role shifts from pilot → mission supervisor.

This creates a major UX challenge:

How do we design interfaces that allow humans to trust and supervise autonomous robots operating in complex environments?

My Role

Product Designer — Autonomy Systems

I worked closely with product managers, robotics engineers, and domain experts to translate autonomy technology into usable workflows.

My work included:

• defining the UX vision for autonomy

• designing mission planning and monitoring workflows

• facilitating cross-team workshops

• aligning UX with robotics architecture and product strategy

• documenting the mission management UX for the autonomy platform
  
  

Understanding the System

The autonomy system is built around four core components:

1. Perception
   Sensors and AI map the environment and estimate the robot’s position. Through SLAM (Simultaneous Localization and Mapping) the system continuously estimates its position and builds a map of the environment.

2. Guidance
   The decision engine that selects behaviours based on mission goals. It executes missions using behaviour trees, selecting actions based on the mission goal and sensor data.

3. Navigation
   Converts mission logic into vehicle movement

4. Vehicle Interface
   Connects the autonomy system to the robot’s control system.

Understanding this architecture allowed me to design interfaces that expose system logic without overwhelming the operator.

The roadmap aims to reach Level 4 supervised autonomy, where robots execute missions while humans supervise and intervene only when necessary.

Process

UX Vision

The UX strategy focused on reducing cognitive load while maintaining human oversight.

Key principles:

Mission-centric UI
Design around mission goals instead of system internals.

Transparency
Operators should always know what the robot is doing and why.

Human-in-the-loop autonomy
Operators supervise missions and can intervene anytime.

Progressive complexity
Simple tasks are easy to start, advanced control remains available.

Designing Mission Management (NDA)

A key part of my work was defining the Mission Management System.

This interface allows operators to:

• plan inspection missions

• monitor robot behaviour

• understand system decisions

• intervene when necessary

The system connects mission planning, execution, monitoring, and post-mission analysis into one workflow.

Mission Planning

Operators define missions using reusable behaviour blocks such as:

• go to waypoint

• follow pipeline

• maintain distance from asset

• inspect structure

• re-inspect if quality is insufficient

These behaviours are composed into larger mission workflows.

This modular structure allows complex missions to be built without coding.

Policy-Driven Autonomy

A key concept I worked with was policy-based mission logic.

Instead of hardcoding behaviour, the system follows policies such as:

• slow down near structures

• maintain a minimum distance to assets

• avoid restricted zones

Policies allow missions to adapt dynamically to the environment while maintaining safety.

Mission Monitoring

Another key UX challenge was making autonomy observable.

Operators must be able to answer questions like:

• Where is the robot?

• What is it doing right now?

• What will it do next?

• Why did it change behaviour?

The monitoring interface therefore exposes:

• robot location and trajectory

• active mission behaviour

• system state and sensor confidence

• alerts and policy triggers

This creates trust and situational awareness, which are critical when supervising autonomous robots.

Impact

The work contributed to the foundation of the EIVA Autonomy System, enabling:

• supervised robotic inspection missions

• perception-driven navigation

• scalable autonomy across subsea robots

The roadmap aims to reach Level 4 supervised autonomy, where robots execute missions while humans supervise and intervene only when necessary.