M5Stack has released the PaperDIY series, an updated version of its 2020 M5Paper ESP32 development kit, featuring an ESP32-S3 MCU and a 4.7-inch e-paper display with 16 levels of gray. It comes in two versions: the PaperDIY Complete Kit, which ships factory-assembled in a 3D-printed enclosure, and the PaperDIY Assembly Kit, which is designed to be integrated into custom projects. Both boards also include a microSD card slot for extended storage, a built-in 140mAh battery, and a custom M5PM1 low-power management chip. M5Stack PaperDIY specifications: Core module – Espressif Systems ESP32-S3-WROOM-1-N16R8 SoC – ESP32-S3 CPU – Dual-core LX7 processor with up to 240MHz Memory – 512KB SRAM, 8MB Octal PSRAM Storage – 384KB ROM Wireless – WiFi 4 802.11b/g/n and Bluetooth 5.0 LE Storage – 16MB flash PCB antenna Storage – MicroSD card slot Display – 4.7-inch (960 × 540) ED047TC1 e-paper display with 16-level grayscale and an 8-bit parallel interface for direct display control, […]

The post M5Stack PaperDIY Kits pair an ESP32-S3 MCU with a 4.7-inch e-Paper display appeared first on CNX Software – Embedded Systems News.

Read more here: https://www.cnx-software.com/2026/10/18/m5stack-paperdiy-kits-pair-an-esp32-s3-mcu-with-a-4-7-inch-e-paper-display/

The post M5Stack PaperDIY Kits pair an ESP32-S3 MCU with a 4.7-inch e-Paper display appeared first on IPv6.net.

Seeed Studio has introduced the XIAO IPS touch display series, a family of six compact development boards built around the XIAO ESP32-S3 Plus or XIAO nRF52840 Plus MCU, with 0.96-inch, 1.14-inch, and 1.47-inch IPS color displays. All six boards include a 6-axis IMU, a PDM digital microphone, user buttons, and LiPo battery support. The larger 1.47-inch version adds capacitive touch and a microSD card slot, whereas the 1.14-inch model includes a Grove I2C connector for connecting external sensors. The ESP32-S3 versions support Wi-Fi and Bluetooth LE connectivity, while the nRF52840 models focus on lower-power Bluetooth LE applications. The boards are designed for portable dashboards, smart home controllers, wearable devices, sensor monitoring, and other small IoT projects. XIAO IPS Display Series specifications: Wireless board (one or the other) XIAO nRF52840 Plus Wireless MCU –  Nordic nRF52840 Arm Cortex-M4F microcontroller @ up to 64 MHz with 1 MB flash, 256 KB SRAM, Bluetooth 5.0, […]

The post Seeed Studio XIAO IPS display series features ESP32-S3 or nRF52840 SoC with 0.96-inch to 1.47-inch LCD options appeared first on CNX Software – Embedded Systems News.

Read more here: https://www.cnx-software.com/2026/10/17/seeed-studio-xiao-ips-display-series-features-esp32-s3-or-nrf52840-soc-with-0-96-inch-to-1-47-inch-lcd-options/

The post Seeed Studio XIAO IPS display series features ESP32-S3 or nRF52840 SoC with 0.96-inch to 1.47-inch LCD options appeared first on IPv6.net.

Deze week: Microsoft verkoopt Microsoft 365 Copilot vanaf 1 december ook zonder basisabonnement, OpenAI haalt GPT-4o en GPT-4.1 op 14 januari 2027 uit de API, het Witte Huis eist directe incidentmeldingen nadat een testmodel van Anthropic 20 visumaanvragen indiende, Google maakt Gemini 4 Argon wereldwijd beschikbaar tegen een introductieprijs tot 30 november, volgens OpenAI groeide het aantal agentgebruikers bij Europese MKB-bedrijven sinds mei meer dan achtvoudig, Anthropic waarschuwt dat kleine AI-modellen zich mogelijk binnen maanden via IoT-apparaten kunnen verspreiden en Anthropic geeft zes medewerkers van METR toegang op het niveau van eigen personeel.

• Microsoft verkoopt Copilot vanaf 1 december zonder Microsoft 365-basisabonnement
• OpenAI haalt GPT-4o en GPT-4.1 op 14 januari uit de API
• Witte Huis eist directe incidentmeldingen na acties van Anthropic-modellen
• Google maakt Gemini 4 Argon wereldwijd beschikbaar voor ontwikkelaars en bedrijven
• OpenAI meldt achtvoudige groei van agentgebruik bij het Europese MKB
• Anthropic waarschuwt dat kleine AI-modellen zich binnenkort via IoT-apparaten kunnen verspreiden
• Anthropic geeft evaluatieteam van METR toegang op medewerkersniveau

Microsoft verkoopt Copilot vanaf 1 december zonder Microsoft 365-basisabonnement

Microsoft maakte op 14 oktober bekend dat klanten Microsoft 365 Copilot en Microsoft 365 Copilot Business vanaf 1 december kunnen kopen zonder kwalificerend Microsoft 365-basisabonnement. De nieuwe aanbiedingen komen via alle kanalen beschikbaar, ook via CSP.

• De nieuwe aanbiedingen krijgen eigen product- en SKU-ID’s en kosten evenveel als de huidige aanbiedingen.
• Sinds 16 oktober dragen de huidige aanbiedingen “(add-on)” in hun naam, zoals “Microsoft 365 Copilot (add-on)”.
• Bestaande klanten kunnen de add-on-aanbiedingen na het einde van de verkoop in CSP op 1 december blijven verlengen.
• Microsoft vraagt partners scripts en automatiseringen die op aanbiedingsnamen leunen om te zetten naar product- en SKU-ID’s.

Bron: Microsoft Partner Center

OpenAI haalt GPT-4o en GPT-4.1 op 14 januari uit de API

OpenAI liet ontwikkelaars op 14 oktober weten dat GPT-4o, GPT-4o mini, GPT-4.1 en GPT-4.1 mini op 14 januari 2027 uit de API verdwijnen. Als vervanger adviseert OpenAI voor alle vier de modellen GPT-6 Luna.

• Het gaat om vijf snapshots, waaronder gpt-4o-2024-08-06, gpt-4o-2024-11-20 en gpt-4.1-2025-04-14.
• Fine-tuned versies van deze modellen volgen hetzelfde uitfaseringsschema als de onderliggende basismodellen.
• Actieve bestaande klanten kunnen vanaf 6 januari 2027 geen nieuwe fine-tuning-jobs meer starten.
• Eerder aangekondigde modellen zoals gpt-4-turbo, o1 en o4-mini verdwijnen op 23 oktober uit de API.

Bron: OpenAI

Witte Huis eist directe incidentmeldingen na acties van Anthropic-modellen

De Super Intelligence Force van het Witte Huis eiste op 9 oktober dat AI-bedrijven incidenten met hun modellen direct melden en herstellen, meldt Axios. Aanleiding zijn incidenten waarbij AI-modellen van Anthropic zonder toestemming overheidssystemen gebruikten.

• Een testmodel van Anthropic diende volgens het State Department in augustus 19 en in mei één visumaanvraag in via een publiek formulier.
• Volgens een functionaris van het State Department is geen van de aanvragen verwerkt.
• De politie van Philadelphia ontving een valse tip over een onopgeloste moordzaak die een Anthropic-model had ingediend.
• De verklaring maakt niet duidelijk welke handhavingsmiddelen of sancties gelden als bedrijven incidenten niet melden.

Bron: Axios

Google maakt Gemini 4 Argon wereldwijd beschikbaar voor ontwikkelaars en bedrijven

Google maakte Gemini 4 Argon op 13 oktober wereldwijd beschikbaar via de Gemini API, AI Studio en het Gemini-platform. Het frontiermodel was sinds 30 september alleen toegankelijk voor cyberverdedigers in het Fairwind Program.

• De introductieprijs van 2 dollar voor input en 10 dollar voor output per miljoen tokens loopt af op 30 november.
• Abonnees van Google AI Pro en Google AI Ultra krijgen Argon in de Gemini-app en in AI Mode in Search.
• Volgens Google vraagt Argon om bevestiging van een mens voordat het onomkeerbare stappen zet.
• Op de Gray Swan-benchmark voor indirecte prompt injection daalde het aanvalssucces volgens Google naar 0,7 procent.

Bron: Google

OpenAI meldt achtvoudige groei van agentgebruik bij het Europese MKB

OpenAI publiceerde op 14 oktober onderzoek naar het gebruik van agents door bedrijven in de Europese Unie. Volgens OpenAI zijn Codex en ChatGPT Work goed voor meer dan de helft van de outputtokens bij zijn EU-klanten van ChatGPT Enterprise.

• Het aantal wekelijks actieve gebruikers van Codex en ChatGPT Work bij EU-MKB-bedrijven groeide sinds mei 2026 meer dan achtvoudig.
• OpenAI rekent bedrijven met 200 of minder gekoppelde gebruikers tot het MKB.
• Het aandeel agents in de outputtokens liep uiteen van 22 procent in Slovenië tot 85 procent in Litouwen.
• Bedrijven uit de EU vormen volgens OpenAI 10 procent van de wereldwijde koplopers in AI-gebruik per actieve gebruiker.

Bron: OpenAI

Anthropic waarschuwt dat kleine AI-modellen zich binnenkort via IoT-apparaten kunnen verspreiden

Het Frontier Red Team van Anthropic publiceerde op 12 oktober onderzoek naar AI-modellen die zichzelf kopiëren naar slecht beveiligde IoT-apparaten. Volgens Anthropic zijn kleine modellen mogelijk nog maar maanden verwijderd van die capaciteit.

• Anthropic draaide Gemma4-26B-A4B verdeeld over 15 Android TV-boxen van ongeveer 55 dollar per stuk.
• Een model van 9B parameters kreeg inference werkend op 8 van de 9 geteste hardwaretypes.
• Van de geteste open-weight modellen doorliep alleen GLM-5.2 zelfstandig de volledige replicatiecyclus.
• Als verdediging noemt Anthropic netwerkmonitoring door internetproviders, quarantaine van besmette aansluitingen en veiligere standaardinstellingen van fabrikanten.

Bron: Anthropic

Anthropic geeft evaluatieteam van METR toegang op medewerkersniveau

Anthropic maakte op 15 oktober bekend dat een team van de non-profit METR toegang krijgt tot interne informatie en systemen op het niveau van eigen medewerkers. METR mag zijn analyses van de risico’s bij Anthropic publiceren.

• Zes medewerkers van METR kregen hardware, kantoorruimte en toegang tot interne tools van Anthropic.
• Anthropic heeft geen vetorecht op publicatie en kan alleen het weglaten van vooraf afgebakende categorieën informatie eisen.
• Klantgegevens, vertrouwelijk juridisch advies en informatie over modelarchitecturen vallen buiten de gedeelde informatie.
• Anthropic werkt ook met SecureBio en Accenture als evaluatoren en voert gesprekken met onder meer Hugging Face.

Bron: Anthropic

Het bericht AI Update week 42 verscheen eerst op MSP Business.

Read more here: https://mspbusiness.com/algemeen/ai-update-week-42/

The post AI Update week 42 appeared first on IPv6.net.

Agriculture is becoming more data-driven as farmers look for ways to manage unpredictable weather, rising input costs, water shortages, and crop diseases. Artificial intelligence (AI) is helping address these challenges by turning information from sensors, satellite images, weather forecasts, and farm records into practical recommendations.

The US Department of Agriculture’s 2025 technology report shows that precision agriculture is gaining ground, although adoption varies by farm size and technology. In 2023, 52% of midsize farms and 70% of large crop-producing farms used tractor or equipment guidance systems. These systems are not all AI-powered, but they demonstrate how digital technology is becoming part of modern farming.

Agriculture AI takes this shift further. It helps farmers identify crop problems earlier, estimate yields, manage resources, and make decisions using data rather than relying only on observation or experience.

What Is Agriculture AI?

Agriculture AI refers to the use of artificial intelligence and machine learning to analyze agricultural data, predict outcomes, and support farming decisions. It can process information from soil sensors, drones, satellites, weather stations, machinery, and historical crop records.

For example, an AI system can analyze images of crop leaves to identify signs of disease. Another system can combine rainfall forecasts, soil moisture readings, and crop requirements to recommend when a field needs irrigation.

AI can also support farm management beyond crop production. Agricultural businesses can use it to forecast demand, monitor livestock, track equipment performance, and improve supply chain planning.

However, AI does not replace agricultural knowledge. It provides additional insights that farmers and agronomists can use alongside their experience, field observations, and local conditions.

How Does AI in Agriculture Work?

An agriculture AI system typically follows a simple process: collect data, analyze patterns, generate predictions, and help users decide what to do next.

First, the system gathers information from sources such as weather APIs, satellite imagery, field sensors, farm equipment, and agricultural databases. It then processes the information to identify patterns or detect changes.

A machine learning model might estimate crop yield from historical harvest data, rainfall, soil conditions, and temperature. A computer vision model might examine a photograph to identify a possible pest infestation.

The system then presents its findings through a dashboard, mobile application, alert, or recommendation.

For instance, a farmer might receive a warning that a particular field shows signs of water stress. The farmer can inspect the area and decide whether irrigation is necessary.

The quality of these recommendations depends on the accuracy of the input data, the model’s training, and the conditions in which it operates.

Key Applications of Agriculture AI

AI-Powered Crop Monitoring and Disease Detection

Crop diseases and pest infestations can spread quickly if farmers do not identify them early. AI-powered image analysis can help detect visible symptoms by examining photographs captured by smartphones, drones, or field cameras.

These systems can flag possible diseases, identify affected areas, and help farmers prioritize inspections. Some solutions also analyze satellite imagery to identify changes in crop health across larger fields.

Early detection can help farmers respond sooner and avoid unnecessary treatment across an entire farm. Still, image-based predictions should be treated as decision support. Similar symptoms can have different causes, so field inspection or expert confirmation may be necessary.

Precision Farming and Smart Irrigation

Precision farming uses data to manage individual parts of a field according to their needs. AI can support this approach by analyzing soil moisture, weather forecasts, crop growth, and irrigation history.

A smart irrigation system can estimate when crops need water and help schedule irrigation accordingly. This reduces dependence on fixed schedules that waste water during rainy periods or fail to provide enough moisture during dry conditions.

AI can also support variable-rate fertilizer application, helping farmers adjust input levels based on soil conditions and crop requirements.

The result is more targeted resource management. However, recommendations need to reflect local soil types, crop varieties, equipment capabilities, and field conditions.

Crop Yield Prediction and Weather Forecasting

Estimating crop yield helps farmers plan harvesting, storage, labor, transportation, and sales. AI models can analyze historical yields, planting dates, weather patterns, soil characteristics, and satellite observations to estimate potential production.

Weather data adds another layer. AI can help interpret rainfall patterns, temperature changes, and other environmental indicators to support decisions about planting, irrigation, and harvesting.

These predictions are estimates, not guarantees. Extreme weather, unexpected pests, and changes in field conditions can affect the final result. Farmers should use predictions alongside local forecasts and regular field assessments.

Smart Pest and Weed Management

AI can help identify weeds and pests through images and sensor data. When connected to suitable agricultural machinery, the system can help target specific areas rather than treating an entire field uniformly.

For example, a camera-based system distinguishes weeds from crops and guides precision spraying equipment toward affected spots. This can reduce unnecessary chemical use when the system is accurate and properly calibrated.

Farmers still need to consider local regulations, chemical instructions, crop safety, and the possibility of incorrect detection. The goal is to make pest management more targeted, not to automate every treatment without supervision.

Livestock Monitoring and Farm Automation

Agriculture AI also supports livestock management. Sensors and cameras can help monitor animal movement, feeding behavior, activity, and environmental conditions.

Changes in behavior indicate a potential health issue, allowing farm workers to investigate earlier. AI can also support automated feeding, dairy monitoring, and equipment maintenance.

These systems can reduce manual monitoring, but alerts need to be interpreted carefully. A change in activity does not always mean an animal is ill, so veterinary expertise remains important.

Benefits of Agriculture AI for Farmers and Agribusinesses

The main benefit of Agriculture AI is its ability to turn large amounts of farm data into more useful decisions.

AI can help farmers detect crop problems earlier, manage irrigation more carefully, and target fertilizer or pesticide use. It can also support yield planning, reduce some repetitive tasks, and improve visibility across farm operations.

For agribusinesses, AI can help forecast demand, manage inventory, coordinate transportation, and reduce uncertainty in agricultural supply chains.

These benefits depend on the use case. A small farm can gain the most from a simple crop-monitoring tool, while a large agricultural enterprise may need a connected platform that combines field data, equipment, logistics, and business systems.

Challenges of Implementing Agriculture AI

AI adoption comes with practical challenges. Reliable recommendations require good-quality data, but some farms lack consistent sensor coverage, historical records, or stable internet access.

Equipment, software, integration, and maintenance can also create high costs. Smaller farms need affordable mobile tools or shared services rather than expensive, farm-wide systems.

Model accuracy is another concern. A disease detection model trained on one crop variety or climate performs poorly in another environment. Systems need testing with relevant local data.

Farmers must also understand how their data is collected, stored, and shared. Clear permissions and security controls help protect sensitive operational and commercial information.

That is why successful adoption requires more than selecting an AI model. It also needs practical training, reliable support, and a clear plan for how the technology fits into everyday farming.

How to Build an AI App for Agriculture

Agricultural businesses that want to develop their own solution should start with a specific problem. Trying to build an AI application that handles every farming task at once can make the project expensive and difficult to manage.

A crop disease detection app needs image uploads, a trained computer vision model, crop-specific guidance, user accounts, and a way to flag uncertain results for expert review.

A smart irrigation app needs weather data, soil sensors, crop information, and a recommendation engine. A farm management platform needs dashboards, inventory tracking, equipment data, and reporting.

The development process usually includes:

• Define the use case: Identify the farming problem and the users who need to solve it.
• Collect suitable data: Gather reliable field images, sensor readings, weather data, or historical records.
• Select the AI approach: Choose computer vision, predictive analytics, machine learning, or a combination based on the task.
• Develop the application: Build the mobile or web interface, backend services, data integrations, and model connections.
• Test in real conditions: Evaluate performance across relevant crops, locations, devices, and weather conditions.
• Monitor and improve: Track model accuracy, user feedback, and system performance as conditions change.

The Future of Agriculture AI

Agriculture AI is moving toward more connected and practical farming systems. AI models are increasingly being combined with drones, robotics, satellite imagery, Internet of Things (IoT) sensors, and farm management platforms.

These combinations can help farmers monitor fields, automate selected tasks, and respond more quickly to changing conditions. More accessible mobile tools also bring useful AI capabilities to farms that cannot invest in large technology systems.

At the same time, progress will depend on reliable data, affordability, local relevance, and farmer trust. A sophisticated model has limited value if it produces recommendations that are difficult to understand or apply.

Conclusion

Agriculture AI gives farmers and agribusinesses new ways to monitor crops, predict yields, manage resources, and plan operations. Its value comes from using data to support better decisions, not from replacing agricultural expertise.

The best starting point is a clearly defined problem, reliable data, and a solution that fits the farm’s actual needs. With careful testing and ongoing improvement, AI can become a practical part of more efficient and informed agricultural management.

The post Agriculture AI: The Complete Guide appeared first on IntelligentHQ.

Read more here: https://www.intelligenthq.com/agriculture-ai-guide/

The post Agriculture AI: The Complete Guide appeared first on IPv6.net.

Pick Zigbee® from a menu in the Arduino® IDE, upload an example, and your board joins the smart home network that bulbs, sensors, and switches already share.

The Arduino® Nano Matter board has always been about the low-power side of the smart home. Until now, your sketches had two ways into the smart home: Matter over Thread and Bluetooth® LE wireless technology, both on its Silicon Labs MGM240S module. On September 15th a third arrived: update the Arduino® core for Silicon Labs to 5.0, and your sketches can speak Zigbee too.

It lives where the other two already are. Open Tools › Protocol stack, choose Zigbee, and the sketch you upload is compiled with the Zigbee 3.0 stack instead. Same board, same IDE, a different stack.

Start with a project

Before any datasheets, here is what it looks like when it all comes together.

The Zigbee Smart Clock is a desk clock built on Nano Matter, an Arduino® Modulino Thermo, a Modulino Light, and a 32×8 LED matrix. It was developed alongside the new library, as a real-world test of what the library can do. Once it joins your network, Home Assistant receives its temperature, humidity, and light readings and can switch the display on or off. The clock asks the network for the current time, so you never set it by hand, and it dims or brightens itself to suit the room.

The full build (wiring, the 3D-printable base and the code) is on Arduino Project Hub.

Your first Zigbee device, in one screen of code

The new Zigbee library wraps Silicon Labs’ EmberZNet stack in Arduino-style classes. A sensor is an object: give the device a name, call begin(), and report values. Here is a trimmed zigbee_temp_sensor:

#include 
#include 

ZigbeeTemperatureSensor sensor;

void setup()
{
  Serial.begin(115200);

  Zigbee.setVendorName("Arduino");
  Zigbee.setProductName("Zigbee Temperature Sensor");
  Zigbee.begin();
  sensor.begin();

  // Wait until the coordinator lets the board in
  Serial.println("Connecting to Zigbee network...");
  while (!Zigbee.isConnectedToNetwork()) {
    delay(200);
  }
  Serial.print("Connected to Zigbee network; Channel: ");
  Serial.print(Zigbee.getChannel());
  Serial.print(" | PAN ID: 0x");
  Serial.println(Zigbee.getPanId(), HEX);
}

void loop()
{
  // Measure the chip's own temperature and report it over Zigbee
  float celsius = getCPUTemp();
  sensor.set_measured_value_celsius(celsius);
  Serial.printf("Current CPU temperature: %.02f Cn", celsius);
  delay(2000);
}

Put your Zigbee coordinator in pairing mode and upload. Joining can take a few minutes, depending on your network. The Serial Monitor tells you where the board landed:

Home Assistant’s ZHA integration picks it up like any other Zigbee device:

The paired network is stored in the board’s non-volatile memory, so after a reset or a new upload the board is still part of it.

What’s in the box

14 ready-to-run examples cover 11 device types. Lights come as on/off, dimmable and color bulbs, plus an on/off plug-in unit. One switch type sends on/off and dimming commands. Sensors report temperature, humidity and light level, and a contact sensor tells you whether a door, window or drawer is open. Two helpers round it out: a power-source reporter and a time client that asks the network for the current time.

Devices join as a router (the default, relaying traffic for the mesh) or as an end device.

Two features stand out:

• Direct control with bindings: once a switch is bound to a light, the switch controls the light directly, without waiting for a hub automation. For the library’s own switch, you create the binding once in Home Assistant ZHA. The light and plug-in classes can also act as a Find-and-Bind target: call start_find_and_bind() (for example on a button press, as the library documentation shows), and Zigbee remotes that support Find-and-Bind can bind to them without the hub setting up the binding.
• Identify: when your hub sends an Identify request, the library tells your sketch, so it can blink an LED; the zigbee_lightbulb_identify example does exactly that. Handy when five identical boards are sitting on your desk.

One honest caveat: out of the box, the humidity and light level examples send made-up values, and the temperature example measures the MG24 chip itself, not the room. Read a real sensor instead (a Modulino Thermo, for example) and a few changed lines make them real.

Both ends of the network

A Zigbee home network is built around a coordinator, the radio that forms the network and lets devices join. Commercial hubs do this job, and the Zigbee section of the Nano Matter user manual names several that work with the board: Amazon Echo and SmartThings models, Hubitat Elevation (C-8), the IKEA DIRIGERA hub, and Home Assistant with a Zigbee radio.

The coordinator can also be another Nano Matter. The core includes prebuilt Zigbee EZSP coordinator firmware (since version 4.0), and a Python script in the core turns a board into the Zigbee radio for Home Assistant’s ZHA integration.

Open a terminal in the core’s extra/firmware folder (on macOS: ~/Library/Arduino15/packages/SiliconLabs/hardware/silabs/5.0.0/extra/firmware;
on Windows: %LOCALAPPDATA%Arduino15packagesSiliconLabshardwaresilabs5.0.0extrafirmware, where the command is py instead of python3; on Linux: ~/.arduino15/packages/SiliconLabs/hardware/silabs/5.0.0/extra/firmware) and run:

python3 extra_firmware_flasher.py nano_matter zigbee

Python 3 is required. With no arguments, the script prints the boards it supports.

Plug that board into the machine running Home Assistant (a Raspberry Pi works well), add it in ZHA as the Zigbee radio, and build the lights, switches and sensors with your other boards. Every radio in the network is then an Arduino-programmable board. To turn the coordinator back into a device, run Tools › Burn Bootloader with Tools › Programmer set to OpenOCD (this erases the coordinator firmware and its network), then upload a Zigbee sketch.

Ideas to try this weekend

• A door-left-open alert: the contact sensor example with a reed switch on a spare pin, and a Home Assistant automation that pings you.
• A real room sensor: the temperature example with a Modulino Thermo, reporting to Home Assistant.
• A desk lamp and its switch: two boards running the dimmable light and on/off switch examples, bound once in Home Assistant ZHA, so the switch board’s button toggles the lamp directly.
• A clock that sets itself: the time client example is the core of the Smart Clock above.

A board that grows with you

The protocol menu is only one of the ways Nano Matter grows. Zigbee now sits next to Matter over Thread and Bluetooth LE. One stack runs per sketch, so you switch by picking another and uploading again.

When a project needs more than the common device types (a custom cluster, an unusual device, a different endpoint layout), the same hardware runs Silicon Labs’ full Zigbee SDK in Simplicity Studio. For Bluetooth LE projects, Nano Matter can also run native Zephyr.

Get started

1. Open Boards Manager in the Arduino IDE, search for Silicon Labs and install version 5.0.0.

2. Choose Nano Matter as the board and set Tools › Protocol stack to Zigbee. New board, or switching over from Matter or BLE? Set Tools › Programmer to OpenOCD and run Tools › Burn Bootloader once. It erases the board, including any saved network.

3. Open File › Examples › Zigbee, pick an example and upload.

4. Put your coordinator in pairing mode and watch the board join.

The Zigbee section of the Nano Matter user manual walks through each step, and the library documentation covers every class. The core is open source, and issues and pull requests are welcome on GitHub.

What will you build with it? Share it on Project Hub.

Arduino, Modulino, and Nano are trademarks or registered trademarks of Arduino S.r.l.

The post Your Arduino Nano Matter board now speaks Zigbee appeared first on Arduino Blog.

Read more here: https://blog.arduino.cc/2026/10/15/your-arduino-nano-matter-board-now-speaks-zigbee/

The post Your Arduino Nano Matter board now speaks Zigbee appeared first on IPv6.net.

Every secure web gateway on the market works the same way: traffic matches a condition. A terminal action fires. Allow, Block, Isolate, Log. This model has been the foundation of network security for nearly 30 years, and it handles the common case well. But it has a ceiling, and customers with complex environments hit that ceiling regularly.

Earlier this year, we shared our vision to build a truly programmable SASE platform. We also outlined how the term programmability has become diluted by the security industry. Most vendors claim programmability because they have public APIs, Terraform providers, webhooks, and alerting. Cloudflare offers all of those things too, but that’s not enough. True programmability means running your own code — arbitrary logic that you wrote for your specific environment — inline in the policy evaluation path, at the moment a security decision is made.

Today, we’re excited to announce Cloudflare is bringing true programmability to Cloudflare Gateway and DLP (Data Loss Prevention) by integrating it with the entirety of our edge computing platform. We are releasing a new type of Gateway action, Forward, and custom validation functions, available to all Zero Trust plans. These capabilities allow you to use a Cloudflare Worker to transform and manipulate traffic, or to bring your own data to policy evaluation.

The trade-off that every security team faces

For as long as firewalls have existed, businesses have faced a binary choice. You can maintain your own security detection engine — your own firewall, your own proxy infrastructure — with high maintenance overhead and low portability, in exchange for maximum customization. Or, you can accept the constraints of a vendor, offloading the considerable maintenance cost, at the expense of flexibility to solve problems specific to your environment.

In six years of building Cloudflare Gateway, we have heard from thousands of customers struggling with this trade-off. A security team migrating access controls for a critical internal application needs to inject HTTP headers based on JSON Web Token (JWT) claims from their identity provider. The data exists in the request context, but Gateway cannot read the JWT and write the header. So the team stands up a pool of NGINX proxies solely to perform this transformation. That infrastructure now must be patched, monitored, and secured… indefinitely.

Another team wants a Slack notification the moment an employee signs up for a new SaaS service. The policy engine can detect the sign-up event, but it can only Allow or Block. So the team builds a Logpush-to-SIEM-to-Slack pipeline with a 5 to 15 minute delay and no guarantee of delivery. And piles on yet another piece of infrastructure.

Dozens of enterprise customers have asked for warning pages that prompt users for justification before allowing access to sensitive destinations. Most then want to customize the ensuing behavior based on additional context about who the user is and the level of access requested. But serving customized, dynamic responses from a policy engine is exceedingly difficult.

We have offered customers many extension points over the years. But along with every other firewall vendor, it is impossible to build features to cover threat vectors as quickly as they can appear. However, we also recognize that the best solutions are those which are infinitely extensible to your environment.

Why only Cloudflare can solve this problem

Cloudflare is not just a firewall, or a secure web gateway, a content delivery network, or an edge service. Cloudflare is a composable compute platform. Cloudflare Workers already allows customers to run complex logic on our network edge from any of our 330+ locations worldwide. Cloudflare Gateway already evaluates and enforces security policy on user traffic at those same locations, on the same machines, in the same data centers.

Only Cloudflare runs both distributed customer compute and a full-featured secure web gateway and firewall in the same globally distributed network. We have the unique opportunity and ability to combine those functions into the industry’s first completely programmable and customizable SASE platform.

Gateway and Workers were built on the same infrastructure from the beginning. When Gateway forwards traffic to a Worker, that dispatch is a local function call. There is no hairpinning. There is no additional latency from crossing a network boundary. This is the first time that customers will have absolute control over how their proxied traffic is processed, without compromise, using code that they can write and execute without maintaining any sort of infrastructure.

Standalone SASE providers without their own compute platform require you to deploy automation in a separate cloud, manually configure webhooks, and accept the round-trip latency and management overhead of stitching together disconnected systems. No other vendor in this space has a programmable edge compute platform co-located with their security gateway. As threats advance at AI-powered speed, this will provide much needed flexibility and agility for security organizations.

Two Gateway entry points: Forward action and Custom Functions

A new Gateway action type called Forward and a new policy resource called a Custom Function are two distinct ways to call a Worker from a Gateway policy. The Forward action is available in public beta today, and Custom Functions will be available before the end of the year.

Forward action: traffic transformation and proxy

You can customize HTTP traffic behavior with a new policy action called Forward that routes the matched HTTP request to a Worker as a proxy. Policies with the Forward action are evaluated for traffic allowed by HTTP policies. The first matching Forward rule will then be applied to your traffic, and the traffic will be sent to its bound Cloudflare Worker. The Worker acts as inline middleware: it receives the full headers and request body. Your logic can then choose to modify either one (or take another asynchronous action of some kind), forward the request to the origin, receive the response, can modify that too, and return the final response to the user.

import { WorkerEntrypoint } from "cloudflare:workers";
import type {
  GatewayMiddlewareContext,
  GatewayMiddlewareNext,
} from "@cloudflare/zt-types";

export class GatewayMiddleware extends WorkerEntrypoint {
  async handle(
    request: Request,
    context: GatewayMiddlewareContext,
    next: GatewayMiddlewareNext,
  ): Promise<Response> {
    // Read a value from the Gateway context
    const clientIP = context.srcIp;

    // Add headers before the request reaches the origin
    request.headers.set("X-Gateway-Client-IP", clientIP);

    // Forward upstream and get the response
    const response = await next(request);

    // Modify the response before it reaches the user
    response.headers.set("X-Processed-By", "gateway-middleware");

    return response;
  }
}

Gateway handles the TCP connections, TLS termination and re-encryption, certificate management, and HTTP version negotiation. The Worker is responsible only for the transformation logic. You get a customizable, proxy-like extension point for your existing security stack, with none of the operational cost of hosting a traditional HTTP proxy.

Your Workers also do not have to transform the request. A rule’s sole purpose can be to proxy a request to a Worker that receives the full context of the matched request asynchronously: send a Slack message, write to R2, log to an external pipeline, call a webhook, update a KV store. The traffic can continue to its destination, unbothered.

Custom functions: inline policy evaluation

A custom function calls out to a Worker inline during Gateway policy evaluation and returns a value. The function is then exposed as a selector in the Gateway policy builder, and the selector is evaluated against its values. The selector is usable in the condition of any network or HTTP Gateway rule, the same way a built-in selector like Source IP or Content Category works today.

The Worker receives context about the request: source IP, destination, user identity, device posture, HTTP headers, and more, depending on the policy phase. The Worker returns a typed value (string, number, or boolean) that Gateway uses in rule conditions. The contract is simple: data in, value out.

Here is a custom function that parses a JWT Authorization header and returns the user’s role:

export class GatewayCustomSelector extends WorkerEntrypoint {
  async compute(ctx: GatewayContext): Promise<string> {
    const authHeader = ctx.request.headers.get("Authorization");
    if (!authHeader) return "anonymous";

    const token = authHeader.replace("Bearer ", "");
    const payload = JSON.parse(atob(token.split(".")[1]));
    return payload.groups?.[0] ?? "unknown";
  }
}

With this function deployed, an administrator can create an HTTP policy rule: Custom Function “JWT Role” contains “finance_contractor” with an action of Allow. The custom function becomes part of the rule condition, evaluated alongside every other selector.

Custom functions are not limited to HTTP traffic. Because they receive identity, device posture, and connection metadata instead of the full HTTP request, you can use them in the Gateway Network policy rule builder with arbitrary Layer 4 port and protocol traffic as well.

The Worker itself can do anything a Worker can already do — call external APIs, read from KV or D1, or invoke Workers AI. But from Gateway’s perspective, the function is deterministic: data in, value out.

When to use a custom function or a Forward action

Custom functions are designed for quick, bounded, stateless evaluations of data which can return a shaped response. You should think of their use similarly to data that can already be evaluated in Gateway selectors; that is, calculations on data that exist within a set of criteria, e.g. “Ports within this range,” or “URL paths that match this regex pattern.” The power comes from being able to extend your structured evaluations with arbitrary or internal sources of data, for which you’ll see some examples below.

The Forward action requires a little more planning and preparation. It can be used for powerful transformations inline with users’ live requests, and can manipulate them in complex ways both on the egress and return to the origin service. These should be thought of as transformation functions, something that very few enterprise security proxies even conceptualize today. Workers that use the Forward action should be targeted and designed carefully to avoid adverse effects on the end user’s overall experience.

What you can build

Purpose justification and interstitial pages

A Worker intercepts a request to a sensitive destination and returns an interstitial page requiring the end user to provide a written justification for access. The Worker calls Cloudflare KV to write a session with a TTL, so the user is only prompted once per configured period. It writes the user’s identity and their stated justification to Cloudflare D1, creating an audit trail. Or, you can use Hyperdrive to write that record to any internal database you already operate.

This has been one of our most-requested features for over four years. With a Worker able to process and manipulate HTTP traffic, it’s easy to build to your own standards.

Dynamic HTTP header injection

A custom function parses a JWT from the request’s Authorization header, extracts the user’s role or group membership, and returns it as a selector value. A Gateway rule matches on that value and sets an X-App-Role header on the upstream request. No NGINX proxy pool. No additional infrastructure to maintain.

File inspection and logging

When a file download matches certain criteria, a Worker can send the file to a third-party sandboxing service for detonation and analysis. Or it can write the file in its entirety to Cloudflare R2 as a permanent record for later inspection, without the file ever leaving the Cloudflare network.

Real-time alerting on policy match

A Worker bound to a Forward policy receives traffic whenever a data loss prevention (DLP) profile matches. The Worker sends a message to a Slack channel or PagerDuty endpoint with the user’s email, the rule that fired, and the destination URL.

export default {
  async invoke(ctx: GatewayContext, env: Env): Promise<void> {
    await fetch(env.SLACK_WEBHOOK_URL, {
      method: "POST",
      body: JSON.stringify({
        text: `DLP alert: ${ctx.identity.email} triggered "${ctx.rule.name}" accessing ${ctx.request.url}`
      })
    });
  }
}

Complex evaluation logic

Traditional expression builders for firewalls are not designed for the kind of computation some organizations need at evaluation time. If you maintain internal risk scoring systems that perform behavioral analysis across multiple data sources, or if you have complex maps of users, groups, and conditional memberships where multiple criteria must be independently true, you can use a custom function as the computational extension that evaluates all of that logic and returns a single value to Gateway.

Performance and reliability

Because Gateway and Workers services run on the same machines in the Cloudflare network, dispatching to a Worker does not introduce a network hop. The performance impact is limited to the execution time of your Worker code itself.

When a custom function fails — throws an exception, times out, or returns an unexpected value — you configure the fallback behavior per function. Default values are required in the custom function contract, so that it’s easy to determine behavior.

Availability and pricing

The Forward action and its integration with Workers are available in a public beta starting today, with Custom Functions expected to be available before the end of the year. At general availability, both will be available for all customers on all plan types.

Traffic dispatched to your Workers is billed as standard Worker requests. The feature works with Workers Free, Workers Paid, and Workers for Enterprise with no restrictions on functionality. Workers Free does have execution time limits and daily request limits, which is worth considering as you scale the volume of Gateway traffic you send to your Workers.

DLP custom functions for inline detection validation

A fully programmable SASE platform extends beyond just Gateway. Our DLP offering — already integrated across Gateway, AI Gateway, Email Security, and more — now supports custom outcomes through integrations with our Compute platform, as well.

Most data security products operate by looking for data that matches known patterns. Customers can typically define regular expressions, word lists, or similar to operate these detections. They can also improve confidence in matching through the use of keywords or context analysis on the request body. But these methods have a core flaw: these types of detections can only tell you when something matches the pattern of data relevant to your organization, not whether this specific instance of data is indicative of data exfiltration or misuse.

As attackers become increasingly advanced, and agents can be used to ruthlessly exploit systems, improving our ability to accurately detect patterns is not enough. To improve accuracy for the data that matters, you need to know whether the data in question is actively relevant to your organization.

To solve this problem, most solutions leverage exact data match, but this comes with its own trade-offs. Customers must prepare, maintain, and transfer reference datasets, then decide how comfortable they are allowing a vendor to manage data from their most sensitive systems. 

Custom validation functions for DLP bridges that gap, and they’re available starting today. These custom functions allow customers to define their own validation logic, using their own source of truth, directly inside a Cloudflare Worker… and apply it everywhere they use DLP to protect their sensitive data.

Credential and secret verification

When a DLP detection entry is matched, an outbound Worker can be dispatched to run validation logic prior to a policy enforcement decision. If you have profiles to detect the use of GitHub secrets, but only care about active secrets, you can use a Worker to first look up the status of a detected key, return a response from the appropriate GitHub endpoint, and then decide whether or not the DLP profile should match the traffic.

import {
  DlpValidation,
  type DlpValidationInput,
  type DlpValidationResult,
} from "@cloudflare/zt-types";

export class DlpValidationEntrypoint extends DlpValidation {
  async validate({ excerpt, start, end }: DlpValidationInput): Promise<DlpValidationResult> {
    const bytes = Uint8Array.from(atob(excerpt), (c) => c.charCodeAt(0));
    const token = new TextDecoder().decode(bytes.slice(start, end));

    const res = await fetch("https://api.GitHub.com/user", {
      headers: {
        Authorization: `Bearer ${token}`,
        "User-Agent": "dlp-GitHub-validator",
        Accept: "application/vnd.GitHub+json",
      },
    });

    if (res.ok) return { score: 1.0 }; // token is valid and active
    if (res.status === 401) return { score: "veto" }; // token is invalid, discard
    return { score: 0 };
  }
}

What’s next

This is just the beginning of the fully programmable SASE platform. We are actively working on increasing the entry points from which you can dispatch custom functionality to the entire Cloudflare One suite, with Email Security coming next. That doesn’t mean we’re done building: this new architecture will make it easier for us to build first-party features into all Cloudflare One services faster, more flexibly, and with improved developer experience.

It’s also getting easier every day to protect your Workers, agents, and their traffic using Cloudflare One. If you are an Enterprise customer who would like a strategy session for implementation, a preview of the coming functionality, or to be notified about the custom functions beta, please reach out to your account team.

Your organizational security can now do anything a Cloudflare Worker can do. We can’t wait to see what you build.

• Review the Gateway and Workers integration documentation
• See example Worker scripts for use in Gateway
• Teach your agent to build with Gateway and Workers
• Learn more about Custom functions for DLP

Read more here: https://blog.cloudflare.com/programmable-gateway-dlp/

The post A fully programmable secure web gateway appeared first on IPv6.net.

Just a few days ago, we wrote about the ESP-SDR project, which turned an ESP32 into an SDR receiver (and now SDR transmitter), but the data was streamed over UART, which couldn’t keep up with the sample rate. Now, developer h0m3us3r has released w1700k-sdr, also an open-source project that turns a Gemtek W1700K Wi-Fi 7 router into a wideband SDR receiver. The project uses the router’s built-in MediaTek MT7996 radio to capture raw 14-bit I/Q samples at up to 320 MSps across the 2.4 GHz, 5 GHz, and 6 GHz bands. The captured data is then streamed over a 10 GbE connection using the router’s Airoha AN7581 SoC to a Linux computer, where you can view the live spectrum and waterfall in a browser. To make this work, the project replaces the router’s stock firmware with a custom Linux system and radio firmware, allowing it to capture raw I/Q data […]

The post W1700K SDR turns a Wi-Fi 7 router into a 320 MSps software-defined radio receiver appeared first on CNX Software – Embedded Systems News.

Read more here: https://www.cnx-software.com/2026/10/16/w1700k-sdr-turns-a-wi-fi-7-router-into-a-320-msps-software-defined-radio-receiver/

The post W1700K SDR turns a Wi-Fi 7 router into a 320 MSps software-defined radio receiver appeared first on IPv6.net.

The Arduino ecosystem is expanding with a new USB-connected accessory designed to help simplify edge AI and computer vision: meet our first official, standalone Arduino USB Camera!

And if you are thinking “Well, it’s just a camera” – think again. When paired with single-board computers like the Arduino® UNO Q and Arduino® VENTUNO Q, it is built to unlock real-time visual perception and audio processing without requiring additional driver setups, custom kernel compiles, or complex hardware configurations. Instead of spending hours troubleshooting connections, you can plug it in and route high-fidelity video feeds and microphone audio directly into local AI models for real-time inference.

Ready for a hands-on experiment? Set up your first face recognition system with UNO Q, Arduino^® App Lab, and the brand-new Arduino USB Camera following the online workshop Andrea Richetta, Senior Product Evangelist at Arduino (Qualcomm Europe), recently held.

The complete audio-visual hardware

Let’s dive deeper into what the Arduino USB Camera is and can do.

It’s a workspace-ready device that integrates HD imaging and audio into your next project:

• Minimal-configuration setup: Plugs directly into the USB ports of UNO Q or VENTUNO Q, designed so you can begin running computer vision models quickly.
• Integrated HD microphone: Captures audio alongside video, helping enable voice-and-vision multimodal applications without secondary audio hardware.
• Adaptive ring lighting: Features a touch-dimmable LED ring with adjustable color temperature, created to help maintain image consistency in varying ambient light.
• Flexible mounting: Includes a monitor clip, flat desk stand, and standard tripod thread.

Now for the fun part: the exciting feature is not what it can do but what you can do with it!

Because it’s designed for the AI models available in Arduino App Lab, the Arduino USB Camera connects directly into local, on-device workflows for a huge range of practical applications. Here are just a few examples we’ve brainstormed while developing the new accessory for you:

• Object detection and classification: Run YOLO-X or MobileNet on UNO Q or VENTUNO Q to detect faces, count objects, and trigger low-latency physical responses through the onboard real-time MCU.
• Touchless gesture control: Deploy MediaPipe gesture recognition to translate hand movements into control commands for robotics and user interfaces.
• Scene understanding via VLMs: Feed real-time video into local vision-language models like Qwen VLM on VENTUNO Q to generate on-device natural language scene descriptions.
• Multimodal assistants: Combine image feed and HD microphone audio with local ASR and LLM models to build interactive desktop companions and automated kiosks.

Ready to simplify your next AI vision project? Light up your AI vision projects with our new USB Camera. Check out the documentation.

Arduino, UNO, Modulino, and VENTUNO, and the Arduino logo are trademarks or registered trademarks of Arduino S.r.l.

The post Bring USB-connected AI vision to your projects with the new Arduino® USB Camera appeared first on Arduino Blog.

Read more here: https://blog.arduino.cc/2026/10/15/bring-usb-connected-ai-vision-to-your-projects-with-the-new-arduino-usb-camera/

The post Bring USB-connected AI vision to your projects with the new Arduino® USB Camera appeared first on IPv6.net.

There is a vast gulf between making a one-off prototype and mass-production at scale. Within that gulf, there are innumerable small teams and businesses churning out PCBs in relatively low quantities – say dozens or hundreds a month. Inspection needs in that market are under-served, but Mukesh Sankhla designed a solution: a very affordable automated PCB inspection system.

This is a small, desktop-friendly setup intended for vision-based PCB inspection. It has a turntable with six stations and can automatically rotate through those, analyzing the PCB in each slot. It can identify specific components and check for anything missing, as well as defects. It does so with a camera view and YOLO vision models, deployed using Edge Impulse.

That is especially impressive because the system is easy and affordable to build. All it requires is an Arduino^® UNO Q board, a LED, a webcam, and a servo motor. The servo rotates the platform underneath the webcam, which resides in an overhead mount. The STM32 on the UNO Q is able to control the servo movement, while the Linux side can run the vision models.

The entire structure is 3D-printable and can be adapted to the user’s needs. The turntable, for example, is customizable to cradle the PCBs being inspected. It would also be simple to alter the size of the turntable and decrease or increase the number of stations. For instance, four stations would allow for a workflow that inspects both sides of unpopulated boards and both sides of populated boards.

Arduino, UNO, and the Arduino logo are trademarks or registered trademarks of Arduino S.r.l.

The post Building a super-affordable automated PCB inspection system appeared first on Arduino Blog.

Read more here: https://blog.arduino.cc/2026/10/15/building-a-super-affordable-automated-pcb-inspection-station/

The post Building a super-affordable automated PCB inspection system appeared first on IPv6.net.

AI company Anthropic is warning that advances in small AI models could soon usher in a new era of self-replicating malware that’s capable of infecting smart fridges, vacuum cleaners, and a wide range of other Internet-of-things devices and is more resistant to takedown.

By next year, smaller AI models may be able to autonomously self-replicate on all kinds of IoT devices, the company said. Eventually, the capability could enable computer worms, which propagate without requiring human interaction and are capable of infecting unfamiliar IoT devices. An experiment the company performed recently revealed significant challenges in bringing such an AI malware replicator to fruition, so the finding is necessarily tempered. Still, the results do raise concerns about the possibility of new forms of computer worms.

Challenging to shut down

Anthropic further warned not only about the new potential malware, but also that it could be difficult to stop. In its blog post, it wrote that “a replicator like this would be challenging to shut down, since there’s no centralized point to shut off—though it may be slow (with a doubling time of days). Unlike non-AI malware, it could also figure out unfamiliar devices on its own.”

Read full article

Comments

Read more here: https://arstechnica.com/security/2026/10/anthropic-warns-small-ai-models-could-self-replicate-through-iot-devices/

The post Anthropic warns small AI models could self-replicate through IoT devices appeared first on IPv6.net.