Embedded Device Observability and OTA

Catch and fix camera faults before they impact customers

Get real-time alerts, crash logs, and performance data to troubleshoot issues remotely. No more time and money wasted on returns or engineering visits for problems that could have been fixed with a patch.

Observe, analyze and update any connected camera, from action cameras to CCTV systems.

Talk to an expert Try the sandbox

Catch and fix camera faults before they impact customers

Without Memfault

Developing connected cameras the old way is inefficient and expensive.

Relying on technical support and unhappy customers to identify failures and faults?

The customer experience left to chance because you don't have visibility into device performance in the field.

Struggling to diagnose issues that are ruining the installer experience and causing churn?

When an issue happens it takes significant time and effort to debug causing frustration and distraction.

Wasting money on RMA's and truck rolls for issues that could be fixed with a patch?

Forced to organize wasteful returns and replacements for issues that later turn out to be software related.

With Memfault

Memfault makes operating connected cameras better for everyone.

Get automatically alerted when your deployed cameras are experiencing problems.

Quality and performance monitoring of every device in your fleet, at all times so you know about problems fast.

Catch issues happening during install and ensure cameras are reliable right out-of-the-box.

Issues automatically prioritized based on impact and associated with crash data ready to investigate.

Minimize wasted spend on returns, site visits and customer churn by fixing issues remotely.

Identify, diagnose, and deploy fixes for issues happening in your fleet, all from one system.

Remotely diagnose faults that happen in the field as easily as if the device was on your desk and deploy OTA updates with full visibility and control.

“With Memfault, I get stack traces and a more streamlined release cycle that is much more reminiscent of web development than embedded development. And that’s fantastic. I can control my launches, I can control my rollouts, and I get telemetry from my entire fleet, which is phenomenal. ”

Mike Ressler | CTO | Diamond Kinetics

“The Latch firmware team uses Memfault on a daily basis: both to proactively monitor for issues in the field and to actively debug issues occurring in our QA and alpha environments. Generally, when a firmware-related issue is reported from the field, we already know about it due to Memfault data and can use Memfault to quickly start digging in to get more context to jumpstart our debug.”

Tyler Wickenhauser | Firmware Manager | Latch

“In the past firing up J-tag and trying to reproduce issues required that the issue be really, really significant before we could justify investigating with a debugger. Now it’s just much easier to start that investigation. It’s going to take an hour instead of a day.”

Chris Merck | VP Engineering | Bond Home

The only Observability platform purpose-built for teams developing hardware products.

Develop connected cameras faster, cheaper and with less stress.

Security Cameras

Your customers count on cameras to be easy to install and reliable from day 1. Get notified automatically if cameras drop offline, collect diagnostics data without needing direct access, and catch problems happening from the moment the camera first powers on.

Doorbell Cameras

Don't let frustrating installs, interoperability issues and excessive battery drain ruin your customers' experience. Get all the data you need to prioritize the biggest impact problems and resolve them fast.

Computer Vision Cameras

Whether you are running AI on the Edge or in the cloud it ads complexity and unpredictability. Make sure your computer vision cameras are stable and reliable with diagnostics and heath data collected during operation in the real world.

Track critical camera reliability metrics fleet-wide and get automatic alerts when things go wrong.

Reboot reason tracking and trace monitoring for crashes and issues
Monitor device reliability metrics like uptime, stability, and connectivity performance
Get automatic alerts to Slack or email when a new issue is identified or performance degrades
Compare performance and stability in different operating states and between software versions

Automatically collect diagnostics data from every device so no more time is wasted reproducing issues locally.

Drill-down from fleet-wide analysis to an individual customer device in seconds
Automatically collect coredumps and logs every time a device experiences a crash or issue
Find and correlate problems in your log data with fast and powerful log search tools
See crash data plus extra contextual data like firmware version, hardware batch, and more

Own your data pipeline end-to-end making fleet-management and monitoring much easier.

Search through all your devices to get updated status information on your fleet
Quickly find and investigate faulty devices based on configurable attributes like site, or serial number
Own your analytics data pipeline so you can get answers without waiting for other teams
Start collecting new metrics with a few lines of code

Deploy firmware updates with confidence and control

Tried and tested OTA update infrastructure proven on millions of devices
Reduce OTA rollout risk with automated staged rollouts
Maintain 100% visibility of device health whilst updates roll out
Mitigate mistakes with strict OTA rollout controls, full audit log and one-button abort

Take a Tour

Will Memfault work for me?

We know embedded devices are all unique, that's why we built Memfault to be super flexible.

My devices aren't always connected - will Memfault work for me?

Yes! Memfault is designed to work on devices with intermittent connectivity and it won’t leave you with gaps in your data. We buffer data on the device when it’s not connected and send it to the cloud once connectivity is available.

My device sends data via a companion device or gateway - will Memfault work for me?

Yes! Memfault works on devices that don’t have direct connectivity. We can send data via more or less any path to cloud, whether that’s via a gateway or a companion device.

Code space on my device is limited - will Memfault work for me?

Almost definitely! Memfault is very flexible and be configured to suit your requirements. The on-device space taken by Memfault varies but standard implementations are as below:

On MCU a standard implementation requires approximately 4.5kb of ROM and 1.5kb of RAM.

On Android Memfault requires less than 40MB of disk space. The installed apks take up the majority of space, at 15-25MB. The remaining data is usually under 15MB.

On Linux Memfault’s Linux Daemon requires around 5MB of disk space and 5MB of RAM

My device contains multiple chips - will Memfault work for me?

Yes! Memfault can collect data from multiple chips on the same device. Talk to us for guidance on the best set-up in these more complicated implementations.

My device is very power constrained - will Memfault drain my battery?

Memfault is deployed to lots of very lower power devices already. Our customers have done rigorous testing and found that Memfault has no measurable impact on battery life.

My device is on a metered network with low bandwidth - will Memfault work for me?

Yes! We have customers sending data over multiple different connectivity types including satellite and various LPWA networks. Memfault can be configured to send data in very small packets (<9 bytes) so works in lots of connectivity constrained use cases.

Why is Memfault better than just collecting logs?

Collecting logs from your devices is great. They can be extremely helpful and Memfault does collect logs alongside other data. But logs can be clunky and time-consuming when trying to root cause issues. They are also not great for tracking overall fleet health. Memfault collects a coredump each time there is a crash or error and combines it with the most recent logs from that event automatically so you get right to the most relevant information when debugging. Memfault also collects metrics from every device in your fleet, giving you an efficient and scalable way to understand overall fleet health.

How does Memfault compare to other observability solutions?

Memfault is the only observability solution purpose-built for devices and not servers and apps. We let you understand not just what’s happening across you fleet and across your firmware versions, we help you understand what’s happening on each individual device. We don’t just tell you how often a particular error or crash is happening, we tell you how many devices have experienced it and what are the common hardware or software characteristics those devices have in common. We then let you drill down to each individual device that has experienced that issue and see exactly what’s happening.

Is Memfault a replacement for a JTAG?

Memfault isn’t exactly a replacement for a JTAG. Memfault is designed to give you the same quality (or better in some cases) data as you would get using a JTAG when the device is on your desk but the device can be deployed anywhere in the world.

Memfault collects coredumps, logs, metrics and more automatically and remotely without requiring any kind of physical access to the device. Once the Memfault SDK is integrated it handles the collection and sending of data from your devices over whatever your chosen connectivity protocol and data pathway is.