This is a ten week project where my team and I built an
AI design for one.
For my team I led this project as the UX Interaction Designer.

Problem Space

Lifeguards have a difficult job of scanning the beach and water looking for any possible problems that might be going on. During a rescue situation, the job can get a lot harder and time is not on a lifeguards side. This design is to resolve the issues that a lifeguard may go through and in doing so it could save time and lives.

Project Description

Soteria is split into a three-part ecosystem. Dispatch, watch, and a drone, all three of these work along one another so everyone is always aware of the situation that is going on.


Haven; in use for dispatch, registers everyone their staff assignments, and equipment for the day.

Haven is able to track, record rescue data, and record live feed footage. This allows dispatch to be most informed during a rescue situation.

Havens weather page is accessed by the main hub along with a log sheet charting all the data for the day from each lifeguard.

Problems I Identified on Haven

Understanding that we needed three systems to work with one another was very difficult. My thought on it was that there needed to be some way of tracking.

- Iris (drone) notifies Atlas (watch) of rescue and then Haven is notified if Atlas (watch) activates the rescue state.

The way Haven interacts to a rescue situation is by getting popups in the corner stating that a lifeguard has activated a rescue situation. When Atlas (watch) gets close to Iris (drone) on a rescue situation Haven would be able to track what's going on along with a live feed video from Iris (drone). This solved how a dispatch lifeguard using Haven was always going to be in the loop of the situation.

Dispatch Wireframe

Sign in and Assignments- AI routes equipment/break schedule.
Hub- Able to access any tower, drone, profile, and data.
Profile View- Displays live or recorded drone footage, recorded past data, guard profile, and equipment assignment.
Weather- Displays a radar, shows time of weather, and allows you to send out weather alerts to watches.
End of Day- AI charts all of the lifeguards information.


IRIS; the AI drone, mainly scans on its own looking for signs of a possible rescue situation.

If a situation comes up then it notifies the lifeguard in charge of IRIS sending a warning popup to ATLAS (watch). From there it is the lifeguard's job to confirm the rescue.

In a rescue state IRIS locks with the watch when it gets close and follows the lifeguard to shore. IRIS goes back to a regular state when the rescue is completed.


ATLAS; watch acts as the main device working between notifying Haven of a situation and being notified from IRIS (drone) about a situation.

The rescue state of ATLAS will need to be unlocked from a squeeze feature, which can be done from any screen of the watch.

The warning popup allows you to click on a live camera view of what the drone is picking up or you’re able to deny the warning state by pulling the screen down.

Problems I Identified on Atlas

Atlas was the most difficult device to create a set of interactions for. The problems that kept showing up was figuring out how to simplify all the rules that a lifeguard has to deal with and implement them into the watch.

It is triggered by the drone detecting a situation that may be going on. The lifeguard is able to view the situation switching back and forth from the drone camera to the drone location.
I wanted to make sure that the AI was never making the decision over human thought which is why it goes into a warning state. Squeezing the watch can send you into the rescue state, which can also be done from any screen. Once in the rescue state, you have all the settings to be tracked by the drone and able to communicate at any time.

The solutions that were figured out are keeping the lifguard's eyes always on the rescue, getting rid of useless weight like a walkie talkie, always being watched/tracked from a third eye (Haven), creating vision from an aerial view, and creating charting for Haven.

Watch Wireflow

AI Identifies potential risks and notifies the guard of any dangerous situations
AI GPS tracks the guard while out on a rescue to ensure the security of both the bather and the guard
Identifies nearest lunch relief and allows for a smooth transition from on and off duty
Keeps track of all interactions the guard makes during the day to make for a simplified daily logging system


Lifeguards have a difficult job scanning the beach and water looking for any possible problems. Soteria links everyone together making sure that safety is always a priority, allowing more time to be put into saving lives. All three systems work seamlessly allowing for the lifeguards to always be aware of the situation even when it is out of reach.

Vision Video