4 Tips To Implement Observability to Ship High-quality, Secure IoT Products

IoT device makers often rely on a reactive development approach, but modern connectivity calls for greater insight into IoT devices in the field. Below are a few tips by FranÃ§ois Baldassari, CEO, Memfault, on implementing device observability to ship high-quality, secure IoT products.

IoT developers face a convergence of demand-side industry shifts. We’re right in the middle of exponential growth in connected devices, from 6 billion in 2016, according to Machina ResearchOpens a new window , to an anticipated 27 billion by 2025, according to IoT AnalyticsOpens a new window . Just as critically, consumer expectations about their connected devices also continue to rise. Whereas developers used to ship a product and never expected to engage with it again, consumers now expect devices to work well, last for a long time, and get regular updates that deliver new and unique features. Broader connectivity also introduces complexity, given the proliferation of platforms and applications and the potential for interoperability issues.Â

Adding to the challenge: IoT developers are expected to meet these demands despite many hardware constraints. Designing for ever-lower footprints, developers must contend with limits around power, memory, bandwidth, CPU cycles, storage, heterogeneous device environment with different CPU architectures, available peripherals, operating systems, and connectivity stacks. The benefit of the doubt that might have been there in the earlier days of the consumer IoT market? Long replaced with outsized expectations.

Traditional approaches to device development cannot meet the challenges of this evolving market. That anachronistic approach saw device developers relying on rigorous QA to surface issues before mass production, reengineering, and debugging what could be detected, then shipping the product with fingers crossed. Eventually, some consumers would encounter unanticipated problems and have to send their device back in for testing and repair, or engineers would be dispatched for on-site fixes in more complicated deployments.

But as connected devices are now central to healthcare, finance, industry, and consumer life, this reactive, time-consuming approach to development isn’t just inadequate and costly; it’s enough to tank a product and inflict long-term brand damage.Â

Observability,by which I mean remote insight into device health and repair capability, offers IoT developers a strategy for monitoring products in the field and pushing updates and patches, often before users are even aware. Implementing a cycle of measuring device health, improving when necessary, and continuously updating creates loyal customers through more secure and better products. Here are four tips to get started.

1. Adopt an Agile Development Approach

Knowing that you can monitor devices once they’ve shipped empowers developers to adopt agile workflows and ship minimum viable products with the expectation of improving via ongoing iteration. Instead of freezing products for months, developers can enter devices into mass production with a Day-0 update approach, where a device’s firmware is frozen in an incomplete state with the plan to update the device in the field.Â

See More: Why You Should Be Shipping Software Updates Like a Cloud Developer

2. Monitor From a Distance

In a competitive device market, relying on a system that essentially relies on consumers to find and report issues is a surefire way to dampen demand. Instead, consider building or implementing diagnostic capabilities that monitor, capture data, and instantly flag issues without requiring any action from the user. Critical metrics like battery life, memory usage, and connectivity state give insight into overall device fleet performance and highlight where there may be potential bugs, release versions that may be affected, and the frequency with which they’re appearing.

3. Plan For Updates and Repairs

When device fleets reach the millions, pushing an update that could negatively impact individual devices carries an undeniable risk. With continuous monitoring, developers can incrementally roll out updates, limiting the impact of new issues to a fraction of the fleet and preventing most users from experiencing multiple back-to-back bug fixes. Alternatively, and if flash memory allows, an A/B update system can download an update in the background with no discernible user impact other than prompting a reboot.Â

4. Secure Your Device

IoT breaches have led to rising consumer demand for more device security. They have also spurred legislative and regulatory attention on creating industry standards, meaning device developers need to know more about ongoing, real-time device operation. Implementing an observability plan can give instant notification of an issue. Developers can then assess the risk level and push updates instantly if necessary. A monitoring system also uncovers root causes, so developers can evaluate if third-party code is to blame and triage it quickly.

See More: UK Cyber Resilience: How the UK is dealing with cyber threats

Conclusion

As the IoT grows, there’s no question that devices will continue to improve and provide ever-greater value for end-users. Building with a plan for observability contributes to this cycle. It gives device makers the tools to circumvent customer frustration and produce a better, more reliable, and more secure product for a loyal and growing customer base.Â

Have you implemented device observability? How has it helped you ship better-quality IoT products? Share with us on FacebookOpens a new window , TwitterOpens a new window , and LinkedInOpens a new window .

MORE ON IOT: