All tracks
TRACK 3 — APP DEVELOPMENT

App Development

Design. Develop. Deploy.

Design and ship mobile apps for Android and iOS. Build with Flutter, React Native, or native stacks — from first sketch to a store-ready deploy.

01SNH-AD-01

ContextOS: Context-Adaptive Mobile Computing

Problem Statement

Smartphones currently behave largely according to predefined settings.

But the optimal device behavior changes continuously according to:

  • Environment
  • User activity
  • Battery state
  • Connectivity
  • Noise
  • Movement
  • Nearby devices
  • Application requirements

Challenge

Build a mobile system that dynamically constructs an execution policy for the device based on real-time context.

For example:

Walking + noisy environment + 12% battery

could automatically change:

  • Notification behavior
  • Sensor sampling
  • Screen activity
  • Network synchronization
  • Audio interaction
  • Background computation

Expected Output

A functional mobile prototype demonstrating:

  • Multi-sensor context collection
  • Context classification
  • Dynamic policy generation
  • Resource allocation
  • Adaptive application behavior
  • Measurable performance improvements
02SNH-AD-02

HapticLang: Universal Haptic Communication Layer

Problem Statement

Mobile devices primarily communicate through:

  • Text
  • Visuals
  • Sound

But touch can communicate much more information than simple vibration patterns.

Build a machine-readable haptic language capable of encoding complex information through adaptive tactile patterns.

Challenge

Create a system that converts structured information into personalized haptic sequences.

The system should consider:

  • Device vibration hardware
  • Duration
  • Intensity
  • Frequency
  • Repetition
  • User learning profile
  • Environmental conditions
  • Urgency

Example

Instead of:

Notification: “Navigation alert.”

The phone communicates:

Direction + urgency + distance + confirmation

through a standardized tactile sequence.

Expected Output

A working mobile prototype demonstrating:

  • Semantic-to-haptic conversion
  • Multiple haptic patterns
  • Device capability adaptation
  • User personalization
  • Haptic feedback decoding
  • Measurable recognition accuracy
03SNH-AD-03

BatteryTwin: Predictive Battery Intelligence

Problem Statement

Battery optimization today is largely reactive.

Mobile systems detect:

“Battery is low.”

But they do not sufficiently reason about:

“What sequence of activities will cause the battery to become insufficient for the user's future objective?”

Challenge

Build a mobile Battery Digital Twin that continuously predicts future energy requirements based on:

  • Application behavior
  • CPU/GPU utilization
  • Network conditions
  • Sensor usage
  • Display activity
  • User routine
  • Charging opportunities
  • Upcoming tasks

The system should predict:

Can the device complete the user's intended workload before the next charging opportunity?

Example

User starts:

“I need navigation + camera + video recording for the next 90 minutes.”

The system estimates future energy consumption and dynamically modifies device behavior to maximize the probability of task completion.

Expected Output

A working prototype demonstrating:

  • Energy consumption monitoring
  • Workload prediction
  • Battery forecasting
  • Future-task simulation
  • Adaptive resource management
  • Comparison against conventional battery management
04SNH-AD-04

DeviceMesh: Phones That Cooperatively Become One Computer

Problem Statement

Millions of smartphones exist around us, each with:

  • CPU
  • GPU
  • Camera
  • Microphone
  • Sensors
  • Storage
  • Network connectivity

Yet mobile applications generally treat each phone as an isolated computing unit.

Challenge

Build a secure system where multiple nearby smartphones can temporarily form a cooperative computing and sensing mesh.

Example

  • Phone AGPU
  • Phone BCamera
  • Phone CMicrophone
  • Phone DNetwork

Together:

Execute one distributed application.

Critical Requirements

The system should handle:

  • Device discovery
  • Capability negotiation
  • Task partitioning
  • Latency
  • Intermittent connectivity
  • Privacy
  • Device failure
  • Workload migration

Expected Output

A functional prototype demonstrating:

  • Multi-device discovery
  • Capability exchange
  • Workload distribution
  • Distributed execution
  • Device failure handling
  • Result aggregation
  • Security controls
05SNH-AD-05

LifeCache: Personal Experience Reconstruction Engine

Problem Statement

People generate enormous amounts of fragmented personal information:

  • Photographs
  • Screenshots
  • Voice notes
  • Locations
  • Calendar events
  • Messages
  • Documents
  • App activity

But these are stored as disconnected files.

The challenge is to transform them into a temporally reconstructed experience graph.

Challenge

Build an on-device system that can reconstruct an event from fragmented personal signals.

Example

A user asks:

“What happened during my college presentation last Thursday?”

The system could reconstruct:

  1. 09:12 — Arrived at campus

  2. 09:40 — Opened presentation

  3. 10:03 — Recorded audio

  4. 10:21 — Captured photograph

  5. 10:45 — Shared document

  6. 11:15 — Left venue

But the system must distinguish:

Direct evidence

from

AI-generated inference.

Expected Output

A working mobile prototype demonstrating:

  • Multimodal data ingestion
  • Temporal event alignment
  • Event clustering
  • Evidence confidence
  • Missing-event inference
  • Contradiction handling
  • Timeline reconstruction