The GENIVI Alliance, which is working toward a scalable software architecture for in-vehicle infotainment (IVI), recently released an IVI software architecture report based on interviews with automakers, tier one firms, semiconductor suppliers, and service providers, including Alliance member and non-member companies.
Relevant IVI architecture suppliers, according to the report, include Microsoft, QNX, MicroItron (favored by Japanese automakers), Android, Linux, and GENIVI.
According to survey respondents, Linux will challenge Microsoft for market share and could become the dominant long- term player in IVI. But respondents also said that if GENIVI is successful, “…most, if not all companies developing Linux for IVI solutions will switch to GENIVI, in part because GENIVI is configured specifically for IVI.”
The stakes are high here. Infotainment is widely considered the hardest automotive electronics application, and survey respondents pegged the average cost of IVI system development at $20-50 million. Software accounts for as much as half of that, respondents believe, but open source development has the potential to cut those costs in half, eventually. Several respondents said the cost of the first GENIVI implementation could be as much or more than using a proprietary stack.
But wait, there’s more, says Strategy Analytics analyst Roger Lanctot, who considers the study incomplete. For example, “Intel’s acquisition of Wind River goes without mention as does the merger of Intel’s Moblin platform with Nokia’s Maemo OS to create Meego,” he writes. Further, Lanctot considers the report’s dim view of QNX’s market prospects unjustified.
In the OS debate, Lanctot notes, Android and Genivi do not line up directly with QNX, Microsoft or Linux. The report makes the first time the GENIVI platform is proposed as a replacement for QNX, Microsoft, or any other OS, indicating a change in strategy for the group.
“Presenting Genivi as a one-for-one substitute for existing real-time operating system solutions is a different proposition from offering a code-sharing/recycling platform intended to reduce development costs,” Lanctot writes. “Obtaining industry buy-in to this vision will take 5-10 years, by which time the market may well have moved on to the next big thing.”