Overland AI
User Interface
Create an interface for navigating and managing off-road autonomous vehicles
The Intention
Overland AI, formed through work at the University of Washington Robot Learning Laboratory and partnerships with DARPA, reached out to us after a Teague Space Industry event to boost their off-road autonomous vehicle UI. Unfamiliar with working with a design team, they signed a six-week engagement in January 2024. Since then, our relationship has flourished, and we continue to expand the application together.
My Role
Senior Interaction Designer
Interaction + UI Design
UI Prototyping
Visual Design
The Full Team
Teague – Seattle
Clint Rule – Creative Director
Amanda Welch – Interaction Designer
David Smith – Principal Visual Designer
Warren Schramm – Technical Director
Jon Rush – Backend Developer
Daniel Swann – Frontend Developer
Ashley Deal – Project Manager
The original interface Overland AI engineers had devleoped.
The Challenge
Build an application that fosters trust while managing AI complexity.
Initially, managing the autonomous vehicle relied on a technical, cost-map-centric app built by engineers. Although real-time data is crucial, this focus didn’t build user trust. We shifted to an interface prioritizing telemetry, tasks, and mapping checkpoints, allowing users more comprehensive insight to the vehicle. “Tasking” was a feature we introduced to create a better interaction model. This new model supports a “set it and forget it” approach, letting the vehicle operate autonomously while alerting users only when attention is needed. This holistic view enhanced usability and confidence in the vehicle’s performance.
My Role
My Role as Interaction Designer
As the Interaction Designer, I collaborated with a small team at Teague, including a creative director, principal visual designer, and technical director. In just a week of brainstorming, we established a foundational strategy that could support future features and technical constraints. I transformed this strategy into wireframes, exploring various interaction behaviors and mechanisms. My process involved iPad sketches with varying fidelity to ensure logical consistency before moving to higher-fidelity designs. While I didn’t establish the visual design, I later contributed to it as the application matured, creating new features and interaction patterns.
Everything should be made as simple as possible, but not simpler.
Understanding AI and Computer Vision
Mapping applications may seem simple, but autonomous vehicles introduce complex challenges. Differentiating between high-autonomy routes and areas requiring user-defined paths was crucial. While computers are often seen as smarter than humans, current computer vision is limited by programming constraints. Humans excel at navigating landscapes, accounting for variables like weather and terrain. To address these limitations, I proposed building at least two types of containers to manage checkpoint tasks:
High Autonomy: Allow the vehicle to choose it’s own path based on situation on the ground.
Low Autonomy: Allow users to define precise travel corridors for the vehicle.
A Growing Lexicon
In complex projects, terminology can easily cause confusion, even among experts. For instance, does “path” refer to a literal route or the vehicle’s journey between points?
During this project, I became certified in Object-Oriented UX (OOUX), which organizes content into “nouns” to bridge the gap between design and development. Using OOUX ideas, I developed a lexicon map to align our team and client, ensuring everyone had a clear understanding of terms when creating solutions.
Wearing Multiple Design Hats
At Teague, we value designers who can think beyond specific roles and tackle what needs to be done holistically. Beginning my career in visual design gave me a deep appreciation for how visuals and animation enhance user experience. Although I didn’t establish the initial visual design for this application, I often seamlessly expanded my role to ensure the interface was both visually and technically optimal. This involved creating multiple design options for team discussion. I love visual design for the elegance and order it can bring to any product or experience.
Devs ♡ Documentation
At Teague, we take pride in delivering work that’s thoroughly documented. While it may not be the most glamorous part of the job, it’s essential for ensuring our work is backed by sound logic and clear visualizations. This can involve breaking down visuals to an atomic level or simply detailing the functionality of every screen we deliver. We also make it a point to document our questions as they arise, keeping everything transparent and well-organized.
The Result:
Working with Overland AI on their autonomous vehicle project was a collaborative journey that extended beyond the typical client-vendor relationship. We embraced their vision as our own, and together transformed complex challenges into user-centric solutions, bridging advanced AI technology with practical, intuitive design. Our partnership with Overland AI remains a testament to the power of collaboration in driving innovation and pushing the boundaries of autonomous technology.
