Creative projects & applications<\/h2>\n\n\n\n
<\/span>Required hardware:<\/strong><\/p>\n\n\n\n Recommended additional accessories:<\/strong><\/p>\n\n\n\n For the first creative project, program the robot arm to grasp the building blocks, which were manufactured using a 3D printer, and stack them in the provided box (figure 1).<\/p>\n\n\n\n To do this, the following steps must first be carried out, as shown in Part 1<\/a><\/strong>:<\/p>\n\n\n\n The programme for stacking components can then be run on the Raspberry Pi.<\/p>\n\n\n\n Below is an excerpt from the code that makes Grab-it perform movements via the Raspberry Pi in conjunction with the Moto Pi.<\/p>\n\n\n\n The next step is to prepare the Grab-it for the Internet of Things (IoT). Similar to a smart home, the robot arm will be controlled via a web interface. The popular Flask web server acts as an interface between the HTML interface and the motor control. This allows the robot arm to be conveniently operated via the web interface.<\/p>\n\n\n\n As well as PC control, HTML control also works via an app and smartphone. When used with a Raspberry Pi, the Grab-it can easily be controlled via a smartphone. The idea behind this is as follows:<\/p>\n\n\n\n First, the system must be prepared for installation. Then, the servo control is implemented in Python, and the Flask web server and HTML interface are created.<\/p>\n\n\n\n 1. Prepare the system:<\/strong><\/p>\n\n\n\n 2. Prepare servo control:<\/strong><\/p>\n\n\n\n 3. Create Flask web server:<\/strong><\/p>\n\n\n\n 4. Create the HTML web interface:<\/strong><\/p>\n\n\n\n 5. JavaScript is also stored:<\/strong><\/p>\n\n\n\n 6. The server can now be started<\/strong> and controlled in the browser. To do this, simply execute this command in the folder with Then open the browser on a PC or smartphone connected to the same Wi-Fi network to access the Raspberry Pi:<\/p>\n\n\n\n The Grab-it can then be easily and conveniently controlled via the web interface on the PC or smartphone, just like a smart home device (figure 6).<\/p>\n\n\n\n You can also choose whether to start the Flask app manually or with Autostart. This is not specified in the code, but in the Raspberry Pi system.<\/p>\n\n\n\n 1. Manual start:<\/strong><\/p>\n\n\n\n 2. Autostart:<\/strong><\/p>\n\n\n\n As well as controlling the robot arm via a smartphone or PC, it is also possible to operate it with a joystick or controller. An Xbox controller was connected to the Raspberry Pi via a USB interface. The arm can then be conveniently controlled via the controller using the Flask server that is already installed.<\/p>\n\n\n\n This allows the Grab-it to be quickly and easily moved to any desired position using the controller.<\/p>\n\n\n\n The following code excerpt shows how to activate the Xbox controller:<\/p>\n\n\n\n If you want to get even more out of your Grab-it, you can significantly improve the robot arm with targeted mechanical and software optimisations.<\/p>\n\n\n\n The Grab-it robot arm can be adapted and optimised with simple mechanical upgrades. Particularly effective upgrades include gripper attachments, extended arms and camera holders, which can be produced with a 3D printer.<\/p>\n\n\n\n Different gripper shapes, such as rubberised tips for round objects or wide jaws for flat parts, improve versatility and precision when gripping. However, extended arm segments increase the working radius and should therefore be supported by more stable servos or lightweight materials, such as carbon, to prevent vibrations.<\/p>\n\n\n\n Mounting a camera above the gripper enables AI vision or colour recognition via OpenCV. To improve stability, replace wobbly plastic connections with metal joints or ball bearings. To connect a camera to the Grab-it, use this code to check the camera is correctly integrated, and then use it for colour or object recognition:<\/p>\n\n\n\n The servo mount can be reinforced or dampened to minimise vibrations. These modifications make the Grab-it more precise, stable and versatile, making it perfect for ambitious maker projects.<\/p>\n\n\n\n Targeted software upgrades make the Grab-it robot arm from JOY-IT significantly smarter and more autonomous. Integrating OpenCV is particularly effective for implementing colour, shape or object recognition using a USB or Raspberry Pi camera.<\/p>\n\n\n\n Simple OpenCV functions enable the robot to analyse live video data, filter colour ranges and determine the positions of objects. When combined with a servo driver, for instance, the robot can automatically approach and grasp recognised objects. Simple filters suffice for colour recognition, while contour recognition or AI-based models such as TensorFlow Lite deliver more precise results.<\/p>\n\n\n\n Example (object recognition with OpenCV):<\/strong><\/p>\n\n\n\n Another upgrade is the addition of motion profiles that react flexibly based on camera data. Additionally, an API or Flask web interface can be used to accept commands from web or smartphone applications. These software enhancements make Grab-it an adaptive system that actively \u2018perceives\u2019 and reacts to tasks.<\/p>\n\n\n\n Although there are only a few dedicated projects on GitHub, the maker community surrounding the JOY-IT Grab-it robot arm provides valuable resources. GitHub<\/a> has a small selection of repositories, including sample sketches for integrating OpenCV or Flask web interfaces.<\/p>\n\n\n\n It is also worth taking a look at platforms such as Thingiverse<\/a> or MyMiniFactory<\/a> for STL files for customised grippers or camera holders. Maker forums and DIY blogs, such as Reddit threads and Instagram stories, offer further insights and tips on topics such as improving the mechanics, fine-tuning the servos, and integrating a camera.<\/p>\n\n\n\n Although there is currently no large, central Grab-it community, you can find individual projects and tutorials by searching for keywords such as ‘JOY-IT Grab-it’. Use these resources as a starting point, adapt them to your setup and document your own development to become part of the growing community.<\/p>\n\n\n\n The second part of the JOY-IT article on Grab-it shows how initial projects can be implemented using the robot arm. Grab-it has also been made ‘IoT-enabled’ and can be controlled via a controller, smartphone or PC.<\/p>\n\n\n\n As the community continues to grow, GitHub and other developer forums offer numerous projects, upgrades and other resources that can be used to expand Grab-it continuously, make it AI-capable and further develop it.<\/p>\n\n\n\n Bilder: Adobe Stock, reichelt elektronik<\/p>\n\n\n\n\n
\n
Creative project \u2013 stacking components<\/h3>\n\n\n\n
![\"An](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/01-bauteile-stapeln-1024x684.webp\")
<\/a>\n
![\"The](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/02-programm-stapeln-1024x557.webp\")
<\/a> \"programs\": {\n \"Wuerfel stapeln\": [\n {\n \"M1\": 1.5,\n \"M2\": 1.35,\n \"M3\": 1.486,\n \"M4\": 1.45,\n \"M5\": 1.5,\n \"M6\": 1.1\n },\n {\n \"M1\": 2.126,\n \"M2\": 1.425,\n \"M3\": 0.942,\n \"M4\": 1.39,\n \"M5\": 2.1,\n \"M6\": 1.1\n },\n {\n \"M1\": 2.123,\n \"M2\": 0.774,\n \"M3\": 0.942,\n \"M4\": 1.122,\n \"M5\": 2.1,\n \"M6\": 1.1\n },\n {\n \"M1\": 2.112,\n \"M2\": 0.786,\n \"M3\": 1.05,\n \"M4\": 1.082,\n \"M5\": 2.1,\n \"M6\": 1.1\n },\n {\n \"M1\": 2.112,\n \"M2\": 0.663,\n \"M3\": 1.05,\n \"M4\": 1.082,\n \"M5\": 2.1,\n \"M6\": 1.1\n },\n {\n \"M1\": 2.112,\n \"M2\": 0.663,\n \"M3\": 1.05,\n \"M4\": 1.082,\n \"M5\": 2.1,\n \"M6\": 0.872\n },\n {\n \"M1\": 2.112,\n \"M2\": 0.876,\n \"M3\": 1.05,\n \"M4\": 1.082,\n \"M5\": 2.1,\n \"M6\": 0.872\n },\n {\n \"M1\": 2.112,\n \"M2\": 1.524,\n \"M3\": 0.71,\n \"M4\": 1.082,\n \"M5\": 1.125,\n \"M6\": 0.872\n },\n<\/code><\/pre>\n\n\n\nGrab-it goes IoT<\/h2>\n\n\n\n
Control via web interface and app<\/h3>\n\n\n\n
\n
\n
sudo raspbi-config\nsudo reboot<\/code><\/pre>\n\n\n\n\n
sudo apt update;\nsudo apt install python3-flask python3-smbus i2c-tools<\/code><\/pre>\n\n\n\n\n
sudo i2cdetect -y 1<\/code><\/pre>\n\n\n\n\n
pca9685_control.py<\/code><\/strong>.\n\n
# pca9685_control.py\nimport smbus\nimport time\nimport math\n\nPCA9685_ADDR = 0x40\nbus = smbus.SMBus(1)\n\nMODE1 = 0x00\nPRESCALE = 0xFE\nLED0_ON_L = 0x06\n\ndef set_pwm_freq(freq_hz=50):\n prescaleval = 25000000.0 \/ (4096.0 * freq_hz) - 1.0\n prescale = int(math.floor(prescaleval + 0.5))\n oldmode = bus.read_byte_data(PCA9685_ADDR, MODE1)\n newmode = (oldmode & 0x7F) | 0x10\n bus.write_byte_data(PCA9685_ADDR, MODE1, newmode)\n bus.write_byte_data(PCA9685_ADDR, PRESCALE, prescale)\n bus.write_byte_data(PCA9685_ADDR, MODE1, oldmode)\n time.sleep(0.005)\n bus.write_byte_data(PCA9685_ADDR, MODE1, oldmode | 0xA1)\n\ndef set_pwm(channel, on, off):\n bus.write_byte_data(PCA9685_ADDR, LED0_ON_L + 4 * channel, on & 0xFF)\n bus.write_byte_data(PCA9685_ADDR, LED0_ON_L + 4 * channel + 1, on >> 8)\n bus.write_byte_data(PCA9685_ADDR, LED0_ON_L + 4 * channel + 2, off & 0xFF)\n bus.write_byte_data(PCA9685_ADDR, LED0_ON_L + 4 * channel + 3, off >> 8)\n\ndef angle_to_pwm(angle):\n pulse_min = 150 # ggf. an deine Servos anpassen\n pulse_max = 600\n return int(pulse_min + (pulse_max - pulse_min) * angle \/ 180)\n\ndef set_servo_angle(channel, angle):\n angle = max(0, min(180, angle))\n pwm_val = angle_to_pwm(angle)\n set_pwm(channel, 0, pwm_val)\n\n# Beispiel-Kan\u00e4le (anpassen falls n\u00f6tig)\nBASE = 0\nSHOULDER = 1\nELBOW = 2\nWRIST = 3\nGRIPPER = 4\n\n# Startposition (optional)\ndef home_position():\n set_servo_angle(BASE, 90)\n set_servo_angle(SHOULDER, 90)\n set_servo_angle(ELBOW, 90)\n set_servo_angle(WRIST, 90)\n set_servo_angle(GRIPPER, 60)\n\n# Beim Import einmal Frequenz setzen\nset_pwm_freq(50)\n<\/code><\/pre>\n\n\n\n\n
app.py<\/code><\/strong> file enables control from Flask (figure 3).<\/li>\n<\/ul>\n\n\n\n![\"The](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/03-flask-webserver-1024x550.webp\")
<\/a>\n
app.py<\/code><\/strong> and create the index.html<\/strong> file inside it (figure 4).<\/li>\n<\/ul>\n\n\n\n![\"This](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/04-web-oberflaeche-1024x550.webp\")
<\/a>\n
![\"Button](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/05-buttons-1024x550.webp\")
<\/a>app.py<\/strong><\/code>:<\/p>\n\n\n\npython3 app.py<\/code><\/code><\/pre>\n\n\n\ne.g.: http:\/\/DEINE_PI_IP:5000<\/code>i.e.: <\/span>http:\/\/192.168.0.23:5000<\/code><\/code><\/pre>\n\n\n\n![\"Example](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/06-weboberflaeche-1024x888.webp\")
<\/a>cd \/home\/pi\/grab-it-web<\/code>\nPython3 app.py<\/code><\/pre>\n\n\n\napp.py Ausf\u00fchrungsrechte geben: sudo chmod +x app.py \n\nServicedatei erstellen: sudo nano \/etc\/systemd\/system\/roboterarm.service\n\nInhalt: \n\n[Unit]\nDescription=Roboterarm Steuerung App\nAfter=network.target\n\n[Service]\nExecStart=\/usr\/bin\/python3 \/home\/pi\/V3\/app.py\nWorkingDirectory=\/home\/pi\/V3\/\nUser=root\n# Setze Restart auf 'always', damit der Service bei einem Absturz neu startet\nRestart=always\n# Optional: Setze eine kleine Verz\u00f6gerung vor dem ersten Neustart\nRestartSec=5\n\n[Install]\nWantedBy=multi-user.target\n\n===================\n\ndaemon-Dienst neu laden: sudo systemctl daemon-reload\nAutostart aktivieren: sudo systemctl enable roboterarm.service\n<\/code><\/pre>\n\n\n\nControl via Xbox Controller<\/h3>\n\n\n\n
\n
python3-evdev<\/strong><\/code>.<\/li>\n\n\n\nXboxController.py<\/strong><\/code> (figure 7).<\/li>\n<\/ol>\n\n\n\n![\"The](\"https:\/\/www.reichelt.com\/magazin\/wp-content\/uploads\/2026\/02\/07-xbox-controller-1024x553.webp\")
<\/a>import threading\nimport time\nfrom evdev import InputDevice, categorize, ecodes, list_devices\n\n# Robotersteuerung importieren - NEUE IMPORT-METHODE\ntry:\n import roboter_arm_steuerung as RAS\n ARM_CONTROLLER = RAS.ARM_CONTROLLER\n MIN_MAX_PULSES = RAS.MIN_MAX_PULSES\nexcept ImportError:\n print(\"FEHLER: roboter_arm_steuerung.py konnte nicht importiert werden.\")\n ARM_CONTROLLER = None\n\n\n# -------------------------------------------------\n# Controller Maps & Einstellungen\n# -------------------------------------------------\n\nSENSITIVITY_FACTOR = 0.005 # Puls\u00e4nderung pro Tick f\u00fcr M1-M4\nSENSITIVITY_FACTOR_M5_M6 = 0.025 # NEU: Puls\u00e4nderung pro Tick f\u00fcr M5-M6 (5x schneller)\nDEADZONE = 0.15 # Stick Drift Filter\n\n# Achsen\nABS_X_LEFT = ecodes.ABS_X\nABS_Y_LEFT = ecodes.ABS_Y\nABS_X_RIGHT = ecodes.ABS_RX\nABS_Y_RIGHT = ecodes.ABS_RY\nABS_LT_TRIGGER = ecodes.ABS_Z\nABS_RT_TRIGGER = ecodes.ABS_RZ\n\n# Buttons\nBTN_START = ecodes.BTN_START\nBTN_BACK = ecodes.BTN_SELECT\nBTN_X = ecodes.BTN_SOUTH\nBTN_Y = ecodes.BTN_NORTH\nBTN_LB = ecodes.BTN_TL # NEU\nBTN_RB = ecodes.BTN_TR # NEU\n\n# Motor-Mapping\nAXIS_MAPPING = {\n ABS_X_LEFT: 'Motor1',\n ABS_Y_LEFT: 'Motor2',\n ABS_Y_RIGHT: 'Motor3',\n ABS_X_RIGHT: 'Motor4',\n ABS_LT_TRIGGER: 'Motor6', # Greifer (\u00d6ffnen)\n ABS_RT_TRIGGER: 'Motor6', # Greifer (Schlie\u00dfen)\n}\n<\/code><\/pre>\n\n\n\nExtensions & modifications<\/h2>\n\n\n\n
Mechanical upgrades for the Grab-it from JOY-IT<\/h3>\n\n\n\n
# camera_test.py\nimport cv2\n\n# 0 = erste Kamera (USB-Webcam oder PiCam \u00fcber \/dev\/video0)\ncap = cv2.VideoCapture(0)\n\nif not cap.isOpened():\n print(\"Kamera konnte nicht ge\u00f6ffnet werden!\")\n exit()\n\nwhile True:\n ret, frame = cap.read()\n if not ret:\n print(\"Kein Kamerabild empfangen!\")\n break\n\n # Bild anzeigen\n cv2.imshow(\"Grab-it Kamera \u2013 Livebild\", frame)\n\n # ESC zum Beenden\n if cv2.waitKey(1) == 27:\n break\n\ncap.release()\ncv2.destroyAllWindows()\n<\/code><\/pre>\n\n\n\nSoftware Upgrades for the Grab-it \u2013 More Intelligence with OpenCV<\/h3>\n\n\n\n
import cv2\n\ncap = cv2.VideoCapture(0)\nwhile True:\n ret, frame = cap.read()\n hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)\n mask = cv2.inRange(hsv, (100,150,0), (140,255,255))\n cv2.imshow(\"Erkennung\", mask)\n if cv2.waitKey(1) == 27:\n break\ncap.release()\ncv2.destroyAllWindows()\n<\/code><\/pre>\n\n\n\nCommunity & resources<\/h3>\n\n\n\n
Grab-it in action and future prospects<\/h2>\n\n\n\n
\n\n\n\n