Khyati Mehra
Google
Organizer
Build with AI · Physical Edition · Kochi 2026
An overnight 24-hour build experience where AI meets hardware. Real sensors. Real models. Real output. Build something that actually does something. Powered by Gemma and Google for Developers.
Until hackathon kickoff · Apr 18, 2026 · 2 PM IST
00
Days
00
Hours
00
Mins
00
Secs
Edge AI
TinyML
Computer Vision
Physical Prototypes
Hardware Hacking
Embedded Systems
Sensors & Actuators
Real-World AI
Build with AI
Kochi 2026
Edge AI
TinyML
Computer Vision
Physical Prototypes
Hardware Hacking
Embedded Systems
Sensors & Actuators
Real-World AI
Build with AI
Kochi 2026
What is this
AI that touches
the real world.
No toy demos. No chatbots. Real hardware, real sensors, real impact.
The Physical AI Hackathon is a curated, in-person build experience designed to bring together developers, makers, and engineers to create AI-powered systems that interact with the physical world.
The event focuses on bridging the gap between software intelligence and real-world applications, encouraging participants to build tangible, working prototypes using AI and hardware.
Physical AI means systems that capture real-world inputs through sensors, process data using AI/ML models (Edge AI, TinyML, Computer Vision), and trigger meaningful actions through physical outputs or devices.
Teams of 2–4 are curated through an application process based on prior experience in hardware, software, or AI/ML. The event begins with a focused session on Edge AI fundamentals.
// Physical AI pipeline
Sense
Cameras, microphones, accelerometers, temp sensors, LiDAR, and more
↓
Process
AI/ML models at the edge — TinyML, CV, NLP on constrained hardware
↓
Act
Motors, displays, alerts, actuators — meaningful output in the real world
Build Themes
Pick your domain.
Open-ended exploration with one hard rule — your project must use AI integrated with physical hardware to solve a real problem.
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
TRACKS·
Sustainability & Climate
AI-powered environmental monitoring, resource optimization, and sensing systems for a
greener planet.
Green
Tech
Accessibility
Hardware + AI solutions that empower people with disabilities and bridge gaps in everyday
life.
Inclusive Tech
Smart Infrastructure
Intelligent systems for cities, buildings, and public spaces — safer, smarter, more
efficient.
Urban AI
Safety & Automation
Real-time detection, anomaly identification, and automated response systems that keep
people safe.
Safety
Agriculture & Rural
Edge AI for crop monitoring, soil analysis, and precision agriculture for rural
communities.
AgriTech
Open Challenge
Your domain, your problem. As long as AI meets hardware and solves something real — build
it.
Open
By the Numbers
The format.
24
Hours of Building
80
Curated Builders
2–4
Team Size
∞
Hardware On-Site
Key Dates
Apr 4
Announcement & Registration Opens
Apr 12
Registration Closes
Apr 13
Selected Participants Notified
Apr 18 · 2 PM
Hackathon Kickoff
Apr 19 · 2 PM
Judging & Prize Distribution
Schedule
24 Hours.
A fully packed schedule from kickoff to demo day — April 18 at 2 PM through April 19 at 4 PM.
Day 1 — April 18
2:00 PM
Event Kickoff & Hackathon Begins ⚡
5:00 PM
Tea Break ☕
7:00 PM
Checkpoint 1
8:30 PM
Dinner 🍽
11:30 PM
Checkpoint 2
Day 2 — April 19
12:30 AM
Midnight Side Events 🌙
8:00 AM
Checkpoint 3
9:00 AM
Breakfast 🍳
12:30 PM
Final Checkpoint + Lunch
2:00 PM
Hackathon Ends & Submissions Due ⚡
2:00 PM
Presentations, Judging & Prize Distribution 🏆
4:00 PM
Event Ends
Judging
What makes a winner.
Projects evaluated on six dimensions. A working prototype counts more than a polished deck.
01
Real-World Relevance
Does it solve a genuine problem? Would someone actually use this?
02
AI + Hardware Integration
How deeply does the AI connect with the physical system?
03
Engineering Complexity
Ambition, technical depth, and sophistication of the build.
04
Functional Execution
It has to work. A demo that runs beats a deck that explains.
05
Originality
Novel approach. Unexpected application. Something genuinely new.
06
Clarity of Presentation
Can you explain what you built, why it matters, and how it works?
Win Something Real
Prizes.
Top teams recieve cash and hardware prizes. Cash awards for best utilization of Gemma 4 and Gemini.
First place
Total value
₹20,000
Cash prize₹10,000
Hardware kit₹10,000
Curated Edge AI / ML components for on-device and edge prototyping.
Position · 1st
Second place
Total value
₹17,000
Cash prize₹10,000
Hardware kit₹7,000
Quality components and tools to keep building after the hackathon.
Position · 2nd
Gemma 4 innovation
Cash award
₹10,000
For the most creative, practical build that showcases Gemma 4 — novel on-device or edge use of open models that fits the Physical AI brief.
Gemma innovation award
Gemini API spotlight
Cash award
₹10,000
For a distinctive use of the Gemini API and API key — clear product sense, responsible design, and a memorable demo.
Gemini API key challenge
What You'll Build
With
Hardware provided.
Shared hardware resources available on-site. Inventory published in advance. Bring your own components too.
On-site inventory
| Item | Category | Qty | Notes |
|---|
Organized & Supported By
Main Program
Hardware & Community Partner
Organizer
Location
TinkerSpace, Kochi.
TinkerSpace, Kalamassery
A 24/7 makerspace in Kalamassery, Kerala — just 100 meters from Seaport-Airport Road.
Refer to the TinkerSpace accessibility & travel guide for step-by-step directions from metro, bus stops, railway stations, and the airport, plus nearby accommodation, food, and essential services.
📍 Kalamassery, Kerala 🚇 Metro Accessible 🛠 24/7 MakerspaceFAQs
Questions?
What is Physical AI?
Physical AI refers to systems that capture real-world inputs through sensors, process data
using AI/ML models (Edge AI, TinyML, Computer Vision), and trigger meaningful actions through physical
outputs or devices. All projects must demonstrate a clear interaction between AI and the physical
environment.
Who can participate?
70-80 curated participants in teams of 2–4, selected through an application process based
on prior experience in hardware, software, or AI/ML. The selection process maintains a balance between
technical depth and collaborative learning. Whether you're strong in hardware or software, teams are
cross-disciplinary.
What hardware and resources will be provided?
Shared hardware resources available on-site including XIAO-ESP32S3 Sense, XIAO nRF52840
Sense, and Grove Vision AI modules. Hardware inventory will be published in advance for planning.
Participants are encouraged to bring additional components. Continuous mentorship support throughout the
event. Hardware partner: MakerGram.
How do I register?
Registration ran from April 4 through April 12, 2026, and is now closed. Selected
participants were notified by April 13th. Teams of 2–4 work together during the 24-hour event starting
April 18th at 2 PM.
Is there a fee to participate?
No — Physical AI Hackathon 2026 is a free event. There is no ticket or registration charge
for selected participants.
Will food be provided?
Yes. Meals and snacks will be provided during the hackathon so you can stay focused on
building. If you have dietary restrictions, tell the organizers when you confirm or as soon as you can so
we can plan ahead.
What are the prerequisites?
Prior experience in at least one of: hardware/embedded systems, software development, or
AI/ML. Teams are ideally cross-disciplinary. The hackathon begins with a focused session introducing Edge AI
fundamentals and approaches to building AI models for constrained hardware environments.
Registration closed
Applications were open through April 12, 2026. Thank you for your interest. Questions about the event? Use the contact links in the footer.
Registration closed · April 12, 2026
Registration closed