104
Autonomous Drone Emergency Response for extreme-heat incidents in Texas.
Designed a heat-emergency response flow that reduces time-to-help through symptom triage, automated dispatch, and live ETA tracking.
OVERVIEW
How can 104 get life-saving aid to someone having a heat emergency when they're far from help?
In six weeks (Jan-Feb 2025), I led UX/UI for this speculative case study alongside branding collaborators. I mapped the end-to-end response from symptom triage through dispatch, ETA visibility, aid delivery, and post-incident guidance.
Research & framing
I grounded the concept in Texas heat risk patterns, user behavior, and response-system constraints before narrowing the solution space.
Prototyping & testing
I moved from wireframes into scenario-based prototypes and tested whether emergency interactions stayed clear under stress.
Iteration & decisions
I revised hierarchy, CTA language, and guidance flows based on observed confusion points and operational constraints.
The Problem
People spending time outdoors-whether hiking, camping, or working-face a high risk of heat exhaustion but often don't recognize the warning signs until it's too late. In remote areas, limited access to water, shade, or medical aid increases risk, especially without proper education-sometimes requiring emergency intervention.
OPPORTUNITY
Closing the gap between triage and a physical handoff in the field
Education apps and emergency calls help, but they do not reliably shrink distance or uncertainty when someone is overheating miles from aid. I framed 104 around one integrated thread: triage, dispatch, visibility, and support actions that continue after delivery.
Reach beyond the road
In trails, jobsites, and event grounds, response delay is often a last-mile issue, not only an awareness issue.
One coherent thread
Users need continuity from “something is wrong” to “what do I do while help is coming.”
Field proof, not just screens
A useful system must support tangible intervention, not just notifications and checklists.
Symptom onset
User: Feels dizzy, fatigued, and unsure if symptoms are serious.
Gap: Delay in escalation and risk recognition.
104: Fast triage prompt with clear urgency language.
Help request
User: Needs to act quickly with low cognitive bandwidth.
Gap: Too many branching choices increase hesitation.
104: Single primary emergency action with guided input.
Dispatch
User: Waits without confidence that support is actually coming.
Gap: No transparent status or ETA visibility.
104: Live dispatch status and drone ETA tracking.
Field handoff
User: Receives supplies but still needs practical direction.
Gap: Resources without instructions can be misused.
104: Contextual guidance tied to delivered supplies.
Recovery
User: Needs safe follow-up after immediate incident.
Gap: Post-incident support is usually disconnected.
104: Aftercare prompts and next-step monitoring.
Scenario Strip
Jordan, a beginner hiker, is two miles into a 102F trail when dizziness and nausea escalate. A friend opens 104, submits symptoms in one guided path, confirms location, and receives a live drone ETA with immediate first-aid instructions while waiting for delivery.
SOLUTION
104: guided heat triage + autonomous supply delivery.
I designed a flow that helps users report symptoms fast, confirms location, and triggers dispatch while guidance remains visible. I chose this structure because uncertainty during wait time was one of the strongest anxiety drivers in early feedback.
CORE FLOWS
These flows show how 104 guides users from first symptom awareness to confirmed support in the field.
Emergency Flow
Emergency home
Symptom check
Confirm location
Track drone ETA
What to do meantime
Delivery handoff
Learn Flow
Course overview
Course details
Lesson content
Watch Flow
.png&w=1920&q=75)
Open app (light mode)
Log hydration
Log success
Open app (default)
Heat alert
Body temperature alert
Hydration reminder
Drone ETA notification
Drone Screens
Drone status screen
Drone detail screen
BACKGROUND & CONTEXT
The critical minutes between symptom escalation and tangible help
In extreme Texas heat, early symptoms often look like normal fatigue until judgment drops. In remote trails, jobsites, and events, distance and connectivity gaps compound that delay, leaving people without clear, immediate support while risk rapidly increases.
104 is designed around this gap: not just heat awareness, but the time between “something feels wrong” and receiving actionable aid on site.
RESEARCH
Understanding where heat emergencies break down
Panel 1 - Survey Snapshot
Panel 2 - Interview Quotes
“I was out of water, dehydrated, hungry, and stopping to rest every few min.”
“I remember walking back into the house and then I was on the ground... I passed out again.”
“I almost passed out from heat exhaustion.”
“Sometimes it is too hot to function. I feel overstimulated by the heat in a way.”
“I begin to overheat, sometimes sending me into a panic attack. I also tend to vomit if I overheat.”
“Overheating tends to make a lot of these symptoms spike up and it is incredibly exhausting and can even ruin the day or trip.”
“I tend to forget to bring certain things with me in the heat...”
Panel 3 - Synthesis
Synthesis: where the drone intervenes
From the survey and interviews, three themes emerged that directly informed the drone concept.
Remote and hard to reach: People often become sick on long hikes, at lakes, or during long walks home, already far from immediate help and sometimes out of water or energy.
Too hot to think clearly: As heat exposure increases, respondents describe feeling overstimulated, panicked, or close to passing out, making it hard to judge risk or self-rescue.
Prepared, yet still at risk: Most people bring some combination of water, snacks, and sunscreen, but still report headaches, nausea, fatigue, anxiety, and ruined outings when conditions change or outings run long.
I combined a 71-response survey, responder interviews, organizer interviews, and secondary sources (CDC, OSHA, NPS, Texas DSHS) to map the full emergency arc from prevention through response.
Key Survey Findings
- Current solutions still lean on passive monitoring instead of active, automated intervention.
- Responders need real-time visibility into dispatch status, payload progress, and field handoff.
- Manual delivery and fixed-station workflows increase response delay in remote heat incidents.
Competitive analysis
| Category | Strengths | Gaps |
|---|---|---|
| OSHA / NIOSH heat safety tools | Science-backed heat index guidance, risk tiers, and preventative recommendations. | Passive by design; no emergency dispatch, no GPS-driven response, and no incident handoff support. |
| On-demand delivery apps | Mature dispatch patterns, real-time tracking, and highly legible ETA communication. | Road-dependent fulfillment and no symptom triage flow for time-critical heat emergencies. |
| Drone logistics platforms | Proven autonomous delivery capability in remote zones where road response is delayed. | Usually B2B or pre-scheduled; not user-triggered with integrated triage and first-aid guidance. |
I analyzed three adjacent product categories: heat safety tools, on-demand delivery products, and drone logistics systems. Each category solved one layer but not the full emergency journey.
- Heat safety apps: reliable education signal, but passive during active incidents.
- On-demand delivery: strong ETA/dispatch UX, weak fit for remote or emergency scenarios.
- Drone logistics: proven field delivery capability, limited user-triggered triage workflows.
User pain points
68%
Experienced severe fatigue or dizziness
Symptoms often escalated before action was taken.
57%
Carried only water
Limited cooling or electrolyte support during heat stress.
41%
Skipped weather checks
Risk awareness dropped before outdoor activity began.
“I thought I'd be fine with one water bottle.”
“I didn't know if Send Help Now meant 911 or drone dispatch.”
I found that preparation behavior looked stronger than emergency readiness: 92% planned water, but many underestimated actual need and delayed escalation once symptoms started.
- 68% reported severe fatigue or dizziness during summer outdoor activity.
- 57% carried only water, with limited cooling or electrolyte support.
- 41% did not check weather conditions before heading out.
“I thought I'd be fine with one water bottle.” and “I didn't know if Send Help Now meant 911 or drone dispatch.” captured both the physical and cognitive sides of risk.
KEY INSIGHTS
What the research made clear
Insight 1: Heat risk is underestimated until judgment drops
Minimal Escalation Timeline
0-20 min
Early discomfort
Mild thirst, fatigue, and heat exposure cues are often ignored.
20-45 min
Cognitive drop
Judgment slips; people delay asking for help.
45-75 min
Acute symptoms
Dizziness, nausea, panic, and instability appear quickly.
75+ min
Crisis risk
Collapse risk increases and self-rescue becomes unlikely.
92% of hikers planned water, but many underestimated actual need. 68% reported severe fatigue or dizziness in summer activity. The design implication was to surface urgency before users fully trust their own self-assessment.
Insight 2: Preparedness breaks at the moment of escalation
Preparedness breaks when heat escalates
% of respondents
Planned for a hot day
80% - Water + sunscreen only
20% - Includes electrolytes / cooling plan
Needed during real heat episodes
30% - Water + rest was enough
70% - Needed cooling, electrolytes, or had to stop
Survey participants typically prepared with a bottle of water and sunscreen. When real heat episodes hit, most reported severe symptoms such as nausea, rapid heart rate, poor coordination, panic, and the need for shade, cooling, electrolytes, or leaving early.
57% carried only water, and many lacked rapid cooling options once symptoms intensified. This shifted the concept away from “better reminders” and toward coordinated delivery of resources users did not already carry.
Insight 3: Operational systems are too fixed for moving incidents
Fixed infrastructure response
- Relies on stationary tents and known checkpoints.
- Coverage weakens as users move farther from aid.
- Escalation often depends on delayed human detection.
Field-responsive drone model
- Dispatch follows the incident, not fixed station proximity.
- Combines triage, ETA visibility, and targeted payload delivery.
- Supports action while EMS is still en route.
Responders prioritized speed, reliability, and integration, while organizers described stationary tents as insufficient in dense environments. I reframed the product as a bridge into existing systems, not a replacement for dispatch.
EXPLORING SOLUTIONS
Directions I explored before converging
Direction A: Education-first prevention
What it was: preparation-heavy alerts and learning tools.
Why not selected: insufficient during active escalation.
Direction B: Dense triage UI
What it was: multi-path symptom and emergency branching.
Why not selected: higher cognitive load and action hesitation in testing.
Direction C: Guided dispatch + response visibility
What it was: one clear primary action, then explicit state visibility.
Why selected: strongest speed/trust balance under stress.
PROTOTYPING AND TESTING
Validation moments that changed the flow
I ran moderated usability sessions (15 participants) using a heat-emergency scenario and think-aloud protocol. 13/15 completed the request flow in under two minutes, but confusion clustered around action hierarchy and post-trigger control.
Observed friction
- 6/15 participants were unsure whether to tap “Report Symptoms” or “SOS.”
- Small symptom controls were hard to use while stressed or shaking.
- Participants asked for clearer post-trigger state and cancellation guardrails.
Representative quotes
- “I didn't know if 'Send Help Now' meant 911 or drone dispatch.” — P3
- “Do I tap Emergency or Report Symptoms?” — P7
- “I kept looking for a cancel button after I hit send.” — P11
DESIGN DECISIONS / CONSTRAINTS
Challenge → decision → outcome across system constraints
Reduce thinking load during escalation
Challenge: participants hesitated in early branching states.
Decision: I simplified CTA hierarchy and exposed one primary next action at a time.
Outcome: request completion remained fast while confusion points became explicit and actionable in later iterations.
Design around regulatory and hardware limits
Challenge: payload, battery range, weather, airspace compliance, and GPS reliability constrain what can be promised in UI.
Decision: I treated constraints as first-class product inputs (coverage assumptions, supply limits, status transparency) instead of burying them in edge states.
Outcome: interaction logic stayed believable and aligned with real-world drone operations.
Fit into existing emergency workflows, not parallel ones
Challenge: responders needed operational visibility and compatibility, not a disconnected tool.
Decision: I prioritized incident state, location confidence, and supply deployment status over dense responder UI controls.
Outcome: the concept positioned 104 as a bridge layer that augments dispatch systems.
Constraint Map
Payload
Commercial drones typically carry 5-10 lbs; our emergency payload target is about 3 lbs.
Range
Typical operating range is around 10-15 miles per charge, shaping coverage and ETA logic.
Weather
High wind above 25 mph, rain, and extreme heat can reduce stability and battery performance.
Airspace
Routes must comply with controlled airspace rules, line-of-sight constraints, and local authorizations.
Connectivity
Remote GPS and weak signal zones require fallback behavior and clear confidence messaging.
VISUAL IDENTITY
Logo and color system
Logo
Color Palette
UI SYSTEM
Interface and component language
OUTCOME / REFLECTION
What this project changed in my practice
13/15
Completed request flow under two minutes
Most participants successfully moved from symptom reporting to dispatch without assistance.
6/15
Early confusion points identified and resolved
CTA hierarchy and action labels were revised after observed hesitation between key actions.
71+
Survey and interview inputs informed priorities
Concept decisions were grounded in behavior patterns from users and responder perspectives.
This case study reinforced that emergency UX is less about interface novelty and more about decision clarity under stress. The most useful shift for me was designing from constraints first: operational reality, not just screen behavior.
What I'd do differently
- Run earlier field-like tests with higher physical stress and environmental noise.
- Prototype offline-first fallback interactions earlier in the concept cycle.
- Validate responder handoff protocols with broader jurisdictional stakeholders sooner.
Next steps
- Validate the concept with additional field responders and park operations teams.
- Stress-test the flow in lower-connectivity and higher-noise scenarios.
- Model regulatory and deployment feasibility with stricter operational assumptions.
