Welcome to Spanos Robotics
This is the beginning of a new generation in educational robotics.
It started with a single idea:
“What if you could build your own intelligent RoBo-friend — right from your desk?”
This is the right place for,,
RoBo(one) , RoBo(radio) , RoBo(mech) , RoBo(eye) and RoBo(chat).
They do many things,
trying to keep You in good mood.
RoBo family is an exciting, innovative educational platform.


At Spanos-Robotics, we believe that the best engineers aren't those with the fastest computers, but those who can achieve the most with the least. Our Robo Family is specifically designed as a high-level educational laboratory. We don't just teach "coding"; we teach Resource Management and Hardware Optimization.
To master a $5 ESP32 is to truly understand the soul of Computing.
The "Architect’s Toolkit": What Students Learn With Us
When a student builds a Spanos-Robot, they transition from a "coder" to a System Specialist. Here is the real-world value of our curriculum:
1. Advanced Memory & Buffer Management
The Benefit: Students learn the "Zero-Footprint" technique.
The Skill: Allocating specific memory blocks for WiFi, variables, and messages. Instead of letting the system crash, students learn to manually define and clear buffers, ensuring the robot remains stable even during heavy processing.
2. Defeating the Watchdog (WDT) & System Timing
The Benefit: Understanding the heartbeat of a CPU.
The Skill: Using yield() and strategic delay() to "feed" the Watchdog Timer. Students learn how to prevent system resets during long AI calculations or complex string searches.
3. Parallel Processing & Multi-CPU Synergy
The Benefit: Learning how to manage "Teams" of processors.
The Skill: Distributing tasks (Vision on CPU #3, Logic on CPU #1, Audio on CPU #2). This teaches students how to handle data synchronization and inter-processor communication—the foundation of modern industrial robotics.
4. Efficient AI Integration & JSON Handling
The Benefit: Connecting local hardware to global intelligence.
The Skill: Students learn to handle Large Language Model (LLM) APIs like Google Gemini. They master the art of "Chunking"—splitting massive JSON responses into manageable segments to bypass memory restrictions.
5. Prompt Engineering as Hardware Control
The Benefit: Language as a programming tool.
The Skill: Crafting precise "System Prompts" that force the AI to respond in a format the microprocessor can understand. This reduces CPU load and increases the speed of real-time interactions.
6. Real-Time Logic & Pattern Recognition
The Benefit: Turning "data" into "memory."
The Skill: Using Voice Activity Detection (VAD) and keyword search within custom-built matrices (Social Memory). Students learn to build scoring algorithms that identify humans based on visual traits, simulating "Social Intelligence."
7. Power & WiFi Optimization
The Benefit: Stability in a wireless world.
The Skill: Managing current load and RF noise. Students learn to strategically disconnect WiFi or lower TxPower to prevent brownouts and interference during critical sensor tasks or servo movements.
8. Peripheral Mastery & Network Control
The Benefit: Making the Robot a "Master of the House."
The Skill: Modifying existing libraries to control diverse components, from thermal printers on a home network to I2S microphones and I2C displays.
The Ultimate Educational Outcome
A student who completes a Spanos-Robotics project doesn't just have a toy; they have a deep technical understanding of:
-Hardware Limitations: How to stay under the RAM ceiling.
-Software Resilience: How to build code that recovers from its own errors.
-Cloud Connectivity: How to bridge a $5 chip with the most powerful AI on Earth.
This is not just robotics. This is the Art of Efficiency.

Meet the RoBo Family – Five Smart Desk Robots
These are not just toys. Each RoBo is a STEM training platform, built with:
Microcontrollers (ESP32, Arduino Mega)
Touch screens and sensors
Speech synthesis and recognition
Robotic arms and compass navigation
Real AI interaction via Gemini 2.5
The most important, is the spirit , they carry, They are made, as if they were Alive !!!!
We tried on making the Robos , to act like persons !
Delays on reply , a 'human' attitude, based on external signals, coding, random selections and finally the speech communication , with a goal, to behave like us,,
Every RoBo Has a Unique Personality and Skillset
💬 RoBo(one)
The Future of Educational Robotics, isn’t just a robot; it’s an intelligent, interactive training platform powered by Google Gemini 2.5 API. Designed to bring AI out of the screen and into the physical world and provide a tactile way to explore cutting-edge technology.
Key Features:
- Intelligence with Personality: Powered by Gemini, RoBo(one) understands complex natural language, providing real-voice responses that make every conversation feel natural.
- Vision-Guided Interaction: Using an ESP32-Cam, RoBo(one) doesn’t just see—it understands. With advanced Image Analysis, it recognizes objects, people, and surroundings in real-time.
- It can detect facial features, gender, age, clothing in people. It asks for the name, remembers it and distinguishes it from others the next time it sees it.
- The "Bring that to me" Game: A built-in educational game that challenges to find and bring specific objects to the robot, Never before the categorization and item recognition, was so fun , playing with friends and Robo.
- Living Memory: RoBo(one) features an "Item Recognition Memory," allowing it to remember things it has seen before and develop a unique "relationship" with its environment.
- The 'QUIZ' game, challenges your general knoledges ! Robo awaits of an answer : 'Yes, that's true' or 'No, that's a lie'
- Smooth Motion & Presence: With a precision-servo head follow-up system, RoBo(one) maintains "eye contact" through motion detection, creating a deep sense of presence.
- Your Personal Assistant: Hear of the ten latest News (with a rotation on seven topics)
, weather forecasts, or the current date and time.
Robo by itself, will speak for the Historical event of current day, also as a truly well-rounded companion, expands the variety to 6 unique categories (Greek Quote / Philosophy, Health / Posture, Word of the Day, Science, Robo-Joke, Mindfulness).
This ensures that even if you hear the briefing every day, it always feels fresh.
It even provides "Status Reports" on its own system health.
- Fully Autonomous & Portable: With onboard power and seamless WiFi connectivity, RoBo(one) is ready to learn anywhere, anytime.
🤖 RoBo(radio)
-FM radio + speaker + antenna
-Memory and reflex games
-Proximity sensor , PIR sensors
-Compass + Bluetooth synchronization with other RoBos
-Saves personal profiles using SD memory
-Talks with you — and with other robots nearby
🤖 RoBo(mech)
-Motorized robotic arm with **6 Degrees of Freedom** (optional)
-Motion memory and self-protection mode
-Compass + Bluetooth synchronization with other RoBos
-Speech synthesis, PIR sensors,
- “Follow Me” function
-Touch screen with built‑in Tic‑Tac‑Toe game
👁️ RoBo(eye)
-Face and object recognition (2 ESP32-CAM modules)
-Intruder detection with 4 PIR sensors
-Dual TFT displays, voice, Bluetooth, compass
-Controls external devices via relay
-Photo capture and SD storage
-Voice interaction
-Compass + Bluetooth synchronization with other RoBos
“No one escapes the vision of RoBo_eye!”
💬 RoBo(chat)
-Powered by Gemini 2.5 with real voice response
-Touch screen with Greek & English support
-SD memory for saving conversations
-Connection to home network's printer for copying Gemini Ai reply.
-Onboard games, , PIR sensors, voice control
-Upgraded mouth animation & led eyes during speech
-Portable and Complete autonomous, on power and connectivity
Play Breakout or chat with RoBo anywhere — even offline!
Robo_one does not need of sensors or buttons,,, It hears of You , It sees of You, and operate Totally Autonomous
Analizes the image and speech with the aid of Gemini Ai , it remembers the objects or persons , full interacts with the humans.
Do short chatting, listen of the news, or even the time, date and the weather info.
Got boring ? Play with friends, the Game ' bring the object ' ! You have 20 seconds , to show of it to Robo_one.
Do you know if the average person swallows 8 spiders a year in their sleep? Answer 'Yes' or 'No' in the QUIZ to find out the truth!!!
Hear the music You prefer on fm radio station, (but Robo_radio will mute it, if wants to speek)
When it detects a new person in the area, asks nine questions to make a person's profile and save the string data to array, so to Remember the person , next time they meet !
Connect the (optional) robotic arm of RoBo_mech and teach it, a repeated work to be done,
Put two RoBos nearby , to start talking among them or with you , if human presence is detected !
or just play tic tac toe Game, Warning !,,Robo does not like loosing !
Meet the amazing RoBo(eye)
It has the ability of vision through two cams !
RoBo(eye) can recognize up to four persons of your family
and Any object with the aid of a Google API
It sees of everyone and everything !
A special project on Robos family ,,is Robo(chat) based on Gemini 2.5 !
Do endless and free of costs chat !
Ask of anything , RoBo(chat) will reply on short , or extra long speech & text
Equipped with autoscroll function and support of Latin and Greek characters
History, Maths, Physic, Technology, programming, Poems, Lyrics on your new song,,,
Do you want to keep the Robo_chat's reply ? No problem,,, Print it on your home printer. It's So simple !
Bored chating ? play a Breakout game, with increased difficulty and speed , (You play, or RoBo can play !)
Available to update to any API model
Like the other members of Robo family, it speaks a lot, with nice face animation, to answer questions, and become Your new Robo-friend.
A whole dictionary of knoledge !
Except English language, can support a second one ,(voice and on screen) , with the aid of personalized characters.
Robo(chat) is precious , in many ways
Each Robo , has a unique personality , with different skills and ambilities.
The educational value of these five Robos is on that ,,,
They are not just toys, They are ed-tech innovation,
RoBo family is an exciting, innovative educational platform,
Built with 20,000+ lines of C++ code, instructions, files and designed for:
-Students and Makers
-STEM educators and robotics clubs
-Online course platforms and developers
-Hardware resellers
Each robot teaches:
* 3D Printing – Print custom parts from STL files
* Electronics Assembly – Build sensors, servos, displays
* C++ Programming (plus comments on any sketch line)– From beginner code to advanced AI
* AI Integration – Use Gemini, Edge Impulse, Google Vision
Here You’re not just running an app.
You’re building a RoBo — your own intelligent companion.
With just ~$50 in components, and your creativity, you can learn modern robotics from the ground up.
Finally, that is the advantage of the Robo - platforms, Open, Expandable, Evolving
Train , Think , Build and make a Robo as is,,,
or Do changes and make your own Unique version of it.
We provide:
Regular code and hardware updates
Open project architecture
Community support
Educational licensing opportunities
In NEXT RoBo generations , we might change any component, but a RoBo will always carry the same spirit.
Michalis Spanos
💬 Follow us on YouTube for weekly videos , updates
https://www.youtube.com/@spanos-robotics/featured
and on Linkedin : https://www.linkedin.com/in/michalis-spanos-042a62400/





The RoBo Family robots engage students like never before!
plus detailed comments in each line , is creating an effective learning method,









How RoBo(one) Thinks: The Triple-Brain System
Most robots have one "brain." RoBo(one) has three. By using three dedicated processors (CPUs) working in perfect harmony, we’ve created a platform that can see, listen, and think all at the same time—without ever slowing down.
1. The Voice & Hearing Center (CPU #2)
The Job: The "Ears" and "Mouth." 👂
How it works: This CPU is dedicated to VAD (Voice Activity Detection). It listens for your voice, filters out background noise, and converts your speech into text. When the robot speaks, this brain ensures the voice is crystal clear.
2. The Vision & Observation Center (CPU #3)
The Job: The "Eyes." 👁️
How it works: Equipped with an ESP32-Cam, this brain is constantly watching for movement. When it sees something, it captures the image and sends it to the AI for analysis. It tells the other brains, "I see a red ball!" or "I see a person smiling!"
3. The Master Logic & AI Center (CPU #1)
The Job: The "Main Brain." 🧠
How it works: This is where the magic happens. It connects to the Gemini 2.5 AI via WiFi to process complex thoughts. It coordinates the other two brains, manages the Item Memory, runs the Onboard Game, and controls the Smooth Motion of the head.
🛠 Why 3 CPUs? (The Educational Advantage)
For our young engineers, this architecture teaches a vital lesson in Parallel Processing:
No Lag: The robot can move its head while it's talking.
Reliability: If one system is busy downloading the weather, the others are still watching and listening.
Real-World Engineering: This is how real-world industrial robots and autonomous cars are built—using distributed intelligence!
RoBo(one) Technical Specifications
Feature Component / Technology Educational Value
Main LogicESP32-S3 (Dual-Core) Learn about high-speed Wi-Fi & I2C logic
Vision SystemESP32-CAM Discover CMOS image sensors & MJPEG processing
Artificial Intelligence Google Gemini 2.5 API Explore Large Language Models (LLM) & Cloud AI
Voice Output I2S DAC + 3W Class D Amp Understand Digital-to-Analog sound conversion
Voice Input High-Sensitivity VAD Mic Learn about sound frequencies & Voice Activation
Head Actuation MG90S Metal Gear Servo Master PWM (Pulse Width Modulation) & Robotics
Connectivity 2.4GHz Wi-Fi / Serial UART Practice machine-to-machine communication
🚀 Next-Gen Evolution: The Detachable Robo Architecture is Here!
This isn't just an upgrade; it’s a complete structural shift toward professional, scalable hardware production. We have successfully designed and implemented a custom Full-Detachable PCB Motherboard System with a modular connector array.
Now, the dual-core MCU brain clicks instantly into the robot's main chassis.
🛠️ The Technical & Manufacturing Advantages:
- Production Speed at Scale: By separating the complex electronic assembly from the mechanical 3D-printed chassis, the time required to build a single Robo has plummeted. This makes mass deployment for large classrooms, schools, and universities incredibly fast and cost-effective.
- Flawless Educational Serviceability: If a student accidentally damages a peripheral or shorts a component during a coding lab, you don’t need to rebuild the entire robot. You simply detach the mother-board in 2 seconds, swap it out, and the robot is back in action.
- Pro-Tier Hardware Esthetics: Moving away from standard point-to-point wiring to a unified, multi-bus PCB layout gives the Robo family the clean, high- end fit and finish of a commercial tech product.
🤖The future of physical AI education is modular, affordable, and incredibly powerful.

🚀 One-Minute Modular Robotics: Redefining How We Build, Learn, and Iterate! 🤖
What if upgrading a robot’s core cognitive behavior was as easy as swapping a cartridge?
In this quick video, I am demonstrating a complete hardware hot-swap on Robo_One. In less than 60 seconds, its entire main brain housing is detached, and a BRAND NEW modular PCB takes its place.
No complex wire tracing, no manual soldering re-checks, and no system downtime. Just pure, optimized modular engineering.
💡 The Architecture Behind the Modularity
By developing a localized connection hub, the frame and body of Robo_One remain an open sandbox canvas, shifting the heavy lifting entirely to the plug-and-play circuit framework. This unlocks significant architectural advantages:
Personality & Code Shifting: Swap from a relaxed verbal assistant blueprint to high-precision sensor-driven state machines instantly, simply by sliding in a different programmed microcontroller profile.
Localized Language Switching: Transition spoken languages, audio streaming buffers, or interaction sub-routines seamlessly between hardware boards.
The Ultimate Academic Sandbox: Students and researchers can explore completely different design prospects—from deep logic routing to multi-CPU communication—without rebuilding mechanical servo structures from scratch. One chassis, infinite cognitive projects.
Streamlined Assembly Timelines: Traditional robotics education gets bogged down by wire fatigue and messy connection bottlenecks. This custom custom routing cuts connection assembly down to zero, shifting focus back to core software iteration and logic design.
Fail fast, upgrade faster, and keep the hardware moving.
Check out the full workflow swap below! 👇



🚀 Dual-Core Dominance: When AI meets Arcade!
A breakthrough in the evolution of RoBo_chat! In these demo, you’re seeing a true feat of hardware orchestration: An ESP32-S3 robot playing Breakout at ultra-fast Level 24 speeds, while simultaneously narrating a story powered by Gemini 3 AI.
Achieving this required more than just code—it required a custom communication architecture between two microprocessors.
🛠️ The Technical "Secret Sauce":
Asynchronous Dual-CPU Orchestration: We split the brain. CPU#1 handles the heavy lifting—Deepgram STT, Gemini AI processing, and Google TTS audio streaming. CPU#2 is the dedicated Performer, managing high-frame-rate game physics and complex TFT face animations.
The "Robo-Watchdog" Sync: To ensure the mouth moves perfectly with the voice without lagging the game, we implemented a hardware-level Sync Pin logic. CPU#1 acts as the conductor, while CPU#2 "peeks" at the sync wire between game frames to stay in perfect rhythm.
Memory Buffer Mastery: Handling large JSON payloads from Gemini on an ESP32-S3 is a challenge. We engineered a Chunk-Processing system to bypass memory buffer restrictions, allowing RoBo_chat to speak long, intelligent replies without crashing the system.
Non-Blocking Animation: By moving away from traditional delays and implementing a state-based animation routine, the robot’s face remains "alive" and expressive even while the CPU is calculating the trajectory of a high-speed ball at Level 24!
🤖 Why this matters:
It’s easy to run AI on a PC. It’s an entirely different challenge to run Real-Time Generative AI, I2S Audio, and High-Speed Gaming on low-cost, low-power microcontrollers. This project proves that with smart memory management and a dedicated hardware-sync protocol, we can bring sophisticated personality to the edge.
🚀 Next-Gen Evolution: The Detachable Robo Architecture is also available on RoBo(chat) !
This isn't just an upgrade; it’s a complete structural shift toward professional, scalable hardware production. We have successfully designed and implemented a custom Full-Detachable PCB Motherboard System with a modular connector array.
Now, the dual-core MCU brain clicks instantly into the robot's main chassis.
Check out the videos to see RoBo_chat in action—playing hard and talking smart!


















