We're excited to share several important platform enhancements designed to improve your engineering productivity and velocity. This month's updates focus on making hardware discovery and configuration easier, along with introducing powerful new features to support your development workflow.
Find it, configure it, run it: New ways to search for supported hardware
Finding the right components for your machine is now faster and more intuitive. Whether you're wiring up a new machine, checking if your existing hardware is supported, or exploring what Viam works with before making a purchase, our enhanced search capabilities have you covered.
Search by hardware in the machine builder
When adding a component to your machine config, you can now search by hardware keywords—like "HC-SR04," "Intel RealSense D435i," or "Orange Pi Zero 2W"—right from the Machine Builder. This search matches your query against several fields across all modules, including the newly added model short descriptions and supported hardware fields.
Add supported hardware to your own modules
Module authors can now specify supported hardware models via the Registry UI. Each model within a module can have different supported hardware—so if you support multiple sensor types, you can be specific about compatibility. We're actively populating this field across all Viam-supported modules and encourage you to do the same for your own modules.
Want to learn more about the new hardware search capabilities? Read our detailed blog post →
Intelligent detection of configuration attributes for your hardware
The new discovery service allows Viam to automatically detect the attribute values needed for your hardware configuration. Instead of manually finding configuration values, the discovery service queries your hardware and returns the appropriate attribute values for your setup.
For example, when setting up a webcam, the "webcam-discovery" model can automatically detect the correct "video_path" value—eliminating the need to run terminal commands to find this information. This simplifies the setup process and reduces configuration errors, especially when dealing with more complex hardware like RTSP camera systems.
Interested in exploring the discovery service for your hardware setup? Check out our complete walkthrough →
Ask AI
Viam wants you to build and deploy smart machines faster—that's why we added the "Ask AI" button. Now, you can get instant, detailed answers about Viam's features, configurations, and best practices without disrupting your workflow. Whether you're unsure about your next step or need help fine-tuning your configuration, Ask AI provides clear, reliable guidance so you can stay focused on what matters. Try it out today and get the support you need, right when you need it!
Machine settings
The Machine Settings config in Viam enhances reliability by standardizing OS updates, network configurations, and logging across your fleet, reducing the risk of errors and downtime. This consistent, hands-off management also frees up teams to focus on more strategic work, boosting overall productivity.
Viam's new Machine Settings give you more control over device provisioning, network management, and operating system settings—all powered by the Viam Agent. Easily configure these settings within the app, with clear documentation alongside to guide you. Viam makes it easy to manage your machines the way you want. Explore Machine Settings today and take full advantage of Viam's flexibility!
LoRaWAN improvements
In January you heard about Viam's support for LoRaWAN sensors, and we're happy to announce several improvements:
- We now provide specific Viam models for popular LoRaWAN sensors (including Dragino temperature and humidity sensor and Dragino water quality sensor). With these, configuration is even simpler: just enter the device EUI and application key (both printed on the sensor's packaging).
- If you're using a different LoRaWAN sensor, you can still use the "generic" sensor model, which now allows you to add the path to the decoder as a link to a GitHub file.
- We added support for sending downlinks to the device, meaning you can configure things like the uplink interval as well as calibrate the sensor directly from the Viam interface.
- On the gateway side, we now provide a specific Viam model for the Waveshare HAT, meaning configuration usually just involves specifying which region you're in (U.S. or Europe).
See the Viam LoRaWAN module GitHub readme for more on all things LoRaWAN!
TensorFlow models
The team has released built-in support in app for training TensorFlow models. Previously we only had built-in support for TFLite models (though you could train a TensorFlow model via custom training scripts).
The TensorFlow model type is best for general-purpose tasks with more computational power, as opposed to TFLite which is best for use on mobile and edge devices with minimal resources.
The option is accessible via the in-app model training page (Dataset page > Train model).
By streamlining hardware discovery, automating configuration, and providing intelligent assistance, we're making it easier than ever to build and deploy smart machines with Viam.
We encourage you to explore these new features (and explore more in our docs) to see how they can enhance your workflow and accelerate your development process.
