Platform engineering isn’t working for embedded software says Qt
Two-thirds of embedded software from organizations with a platform engineering strategy is still created with custom, ad hoc tools says a study commissioned by the Qt Group in Finland.
Research from Forrester Consulting commissioned by the Qt Group found that existing platforms are not yet meeting expectations and the need to create bespoke software hits development productivity. The online survey covered 317 respondents across Europe and North America and evenly split between automotive, consumer, medical and indistrial.
Almost two-thirds (63%) of embedded code from development teams with a platform engineering strategy is still written with custom tools. Working across devices and OS environments is a top challenge for more than half (51%) of embedded teams with a platform engineering strategy.
All this is despite the respondents’ perceived maturity of their organization’s platform engineering strategy: 65% see their platform as the foundation upon which embedded software is built and 93% of leadership supporting existing platform engineering strategies. But Forrester does not highlight the platform systems that are being used.
- Qt extends Qualcomm deal for industrial IoT graphics
- MCUs add graphics framework for intelligent rendering
Half of embedded developers with a platform engineering strategy (49%) struggle to balance reusability of standardized, high-quality components with the need to adapt platforms to various use cases, hardware, and software solutions. The specific needs of software and hardware platforms, combined with the need to integrate a variety of non-standardized devices and functionalities, still impose a large amount of custom development on teams, preventing them from maximizing the value from their platforms.
The platforms fall down in two areas, supporting configuration-based development and a platform development mindset.
Despite the perceived maturity, that 63% of embedded software is created via custom, project-based development rather than by configuring predefined building blocks. The need to create bespoke solutions for varying products and customer needs indicates that the level of standardization achieved so far is still low and the variety of use cases still poses a major challenge.
While 68% track platform feature usage and 65% track developer productivity metrics, only 38% of respondents track developer experience metrics. Since developers are the platform customers, such a low score indicates that most companies’ platform-as-a-product culture is still far from embedded.
Computing resource constraints, such as memory, energy consumption, and CPU performance, affected 38% of respondents. Moreover, 40% of respondents reported their organization struggles with managing updates and iterations of the platform while minimizing the impact on embedded software development. These constraints force teams to allocate resources to manage these limitations instead of focusing on development.
Machine-to-machine communication and sensor integration is the key differentiating factor. With the trend towards IoT, machine-to-machine communication empowers devices to become interconnected, making it a top differentiating component (46%). Sensor integration — enabling data collection — ranked as the second highest differentiator (44%). Given the strong device focus of embedded developers, cloud connectivity ranked lower than machine-to-machine and sensor, while still 38% considered it a differentiator.
Flexible development through customizable APIs. Allowing developers to tailor third-party integration to the specific needs of the system, customizable APIs ranked as the third differentiator (44%). This also corresponds to the difficulty of integrating third-party APIs — one of the top challenges across the development cycle.
Compliance with industry standards and regulations and security and encryption features are critical, especially given the increased number of cyberthreats and new European legislation and 34% considered them differentiators.
However the report found that the primary reason to use platform tools is to improve end-customer experience (68%) with more reliable, higher-quality reliable software and enhance compliance with industry standards and security (56%). Then it is improved brand identity (57%) and workflow efficiency (54%) from streamlined workflows that help maintain a consistent look and functionality across products.
“We have noticed a gap in the market between the perceived maturity of platform engineering strategies and actual benefits derived from them. Too much work is still being done manually to address the specific needs, whether it be in sectors like medtech, automotive, or industrial automation,” says Juhapekka Niemi, Executive Vice President at Qt Group.
“Platforms should be designed to support change, integrate with evolving technology, and work across a variety of hardware and software platforms. Leveraging a flexible, scalable, and quality-assured framework is key – as is using optimized cross-platform components that are easily deployable. This has been Qt’s guiding principle for many years, and it’s why we’ve seen many customers leverage the Qt Framework as the cornerstone of their platform strategy.”
Because embedded development requires diverse specialized skills, 50% of respondents say talent shortages are the top blockers to better platform strategies. Other key blockers include difficulty integrating legacy platforms (49%) and cultural resistance from product teams (34%).
“On the one hand, embedded engineers face the challenge of understanding the end-user experience for the technical machines they design,” says Maurice Kalinowski, Product Director at Qt Group. “On the other, more widely available web developers, who focus on usability, face steep learning curves in adapting to the embedded space and its intricacies like C/C++ plug-ins, cross platform development, etc. This makes the need for unifying standards stronger than ever.”
End to end tools are not seen as the solution. Just a third (35%) prioritize all-in-one, end-to-end capabilities. Integration is important to respondents in situations where, for example, specialized security tools need to be compatible within a development framework to ensure robust protection without compromising on performance.
As a result half of respondents (49%) see safety, security, and compliance as an essential outcome of their platform engineering strategy. Around 43% also emphasized the importance of customizable APIs for third-party integrations.
“Different APIs have varying agreements and latency times, exposing systems to more security risks and privacy regulations, as developers need to ensure the third-party software handles sensitive data appropriately,” says Miao Luo, Director of Technology Strategy. “The best way for platform teams to mitigate this is by closely monitoring technical debt and minimizing the number of third-party elements in play.”
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
