Riding the embedded software wave

By Anand Sundaram, Senior Associate for PwC’s Operations Consulting

Software that controls and powers embedded devices is playing a key role in making possible the highly integrated, multi-functional ‘smart’ devices we take for granted in our daily lives – from the ubiquitous smart phones/tablet to ‘smart’ home appliances and wearable electronics.

F1

Smart Connected Device Unit shipments by Product Category, 2012-2017 (in millions). Source: IDC’s Worldwide Smart Connected Device Tracker Forecast Report, May 2013. Totals may not equal 100% due to rounding.

Despite the wide variety of features and attributes, a few key trends stand out that are common to most  devices:

  • Explosion in device functionality: Many of today’s highly-integrated devices have impressive communication, computing and sensing capabilities, resulting in the convergence of many functions on a single device.
  • Increased focus on customer experience: Today’s embedded devices are not only expected to deliver a rich set of features and performance at an acceptable price point, but are also expected to be aesthetically pleasing and easy-to-use – tangible attributes that directly drive a customer’s perception of value.
  • Increasing speed of innovation: In a dynamic market environment where products are being launched every 12-18 months, companies are trying to keep up with faster time-to-market expectations and driving an increased pace of innovation.

This results in intense competition, constant product category reinvention, and price/cost pressures for OEMs (Original Equipment Manufacturers). However, there is also an impact on the broader value chain, including device makers that design and sell semiconductor chips for these embedded applications. Embedded device OEMs, under tremendous pressure to improve their time-to-market and reduce costs, have begun requiring semiconductor companies to provide not just a chip, but a complete system solution. This is vastly different from the reality of the decades past (figure 2b), where OEMs played the role of system software integrators, requiring System-on-a-Chip (SoC) vendors to provide just the basic, lower-level building blocks.

F2

Embedded Software Model and Work-Split

What does this increased responsibility for embedded software mean for semi device makers?  Should chip makers reluctantly enter this uncharted territory, viewing it as nothing more than a “new normal” in hyper-competitive markets? Is it possible to strategically reposition their overall value proposition to build a long-term sustainable competitive advantage? Does the current operating model enable chip companies to deal with the added complexity of delivering quality embedded software? How efficient is the R&D organization in delivering to the new paradigm? As we try to find answers to these questions, let us examine the opportunities and discuss what companies should do in order to address the unique challenges that lie ahead.

Benefiting from the changing landscape

These trends have the potential to open new opportunities for semiconductor device makers, just by virtue of their position as key enablers of the value chain. Here are some ways in which companies can build a competitive advantage over rivals, if they play their cards right. Many leading companies are slowly recognizing the underlying opportunities:

  • Enhanced product differentiation: Device makers could differentiate themselves by moving to a solution offering that includes a complete library of embedded software modules – feature-rich middleware such as protocol stacks, integrated graphical user interfaces, and integrated security software. Also, they could use embedded software to deliver tremendous flexibility to incorporate evolving technology standards (e.g., communication protocols) on the same silicon device with a minor software update. Thirdly, companies could target the same device to multiple applications by using embedded software to “tune” the chip performance, thus allowing OEM customers to quickly create product variants for minor, incremental effort.
  • Increased platform leverage: Companies could employ advanced architecture exploration as part of the design process to arrive at intelligent “hardware – software” partitioning, to result in a single hardware device that can be customized to the needs of different applications or customers. Creating “common/unified” hardware, and building the differentiating functionality in software increases platform leverage (fewer distinct silicon devices), and hence improves the return on R&D investment. This strategy not only reduces the cost of designing and managing multiple silicon versions, but also simplifies the overall supply chain by reducing operational complexity.
  • New revenue streams: Although there is already a growing perception that the “value add” has increasingly come to reside in the software, much of today’s embedded software is given away for free in order to secure the “design wins”. However, opportunities to monetize it could arise if companies make the right strategic moves to hedge against a future, wherein hardware is a commodity “vehicle” to deliver value-added functionality to the end consumer. Bundled middleware and application software could be sold as packaged software or on a one-time NRE (Non-Recurring Engineering) fee followed licence royalty model for every unit sold. Significant upgrades to functionality could also be software enabled and delivered to customers remotely, for a fee.

How can companies respond?

Irrespective of where companies find themselves on this journey from chip to solution companies, it is worthwhile to take a holistic view of strategy through execution. PwC’s Product Innovation and Development team has identified a few areas of focus for device companies to help position themselves  for success:

Revisit the overall strategy:  Firstly, companies should rethink their overall strategy, focused on:

  • Their role in the evolving ecosystem and the types of embedded software to offer, e.g., system IP, firmware, and development tools, etc.
  • The markets/applications to offer system solutions, supported by customer value proposition and business case
  • An operational assessment to determine capability gaps, and how to address them, e.g., organically, via partnerships, or through M&A

The clamour of M&A activity indicates that chip vendors are acquiring software companies to rapidly gain the required system knowledge and incorporate it into their “complete solution” offerings. PwC’s analysis indicated that  between January 2010 and June 2013, semiconductor firms acquired 16 companies for system software, 3 for application software, and 8 for the software development tools. There were a total of 34 M&A transactions by semiconductor firms during this period (including private placements)

Design a new operating model: Next, companies should redesign their operating model in alignment with their strategy, focused on the following key areas:

F3

  • Agile development methodology: Companies should consider adopting the Agile methodology or its variants to helping them focus on quality, and flexibly respond to ever changing customer requirements effectively. For a detailed look at effective Agile adoption, read PwC’s whitepaper titled Accelerating embedded software development via agile techniques: The nine strategies that lead to successful embedded software development
  • Organizational structure: Companies should work to build a collaborative model of development that encompasses all ecosystem partners and OEM customers build teams. Cross-functional teams with broad application and system know-how will allow teams to make intelligent trade-off decisions that best suits the market or application. Teams should be accountable for overall solution, and empowered to drive rapid decision-making
  • Partnership models: Operationalizing the partnership model is essential to drive go-to-market efforts, and is achieved through a collaborative, cross-enterprise level effort that spans R&D, marketing, sales and customer support
  • Software productivity tools: Companies should evaluate and adopt new development tools that can help improve R&D productivity by speeding up the design and testing process. The increased sophistication of hardware-software co-development requires tools such as virtual prototyping, IDEs, and emulators that enable fast and effective testing of complex, deeply embedded hardware-software functionality. Specialized Requirements Management tools can also help effectively manage complex hardware/software dependencies, thereby delivering improved efficiency and quality

Conclusion

As companies embark on this operational transformation, it is important to periodically benchmark performance against leading companies. Metrics should account for effectiveness of R&D investments, development productivity, speed of execution, and quality. The new paradigm demands new solutions.

For more information contact Anand Sundaram at [email protected]

POST A COMMENT

Easily post a comment below using your Linkedin, Twitter, Google or Facebook account. Comments won't automatically be posted to your social media accounts unless you select to share.