The automotive industry is going through tumultuous times. The megatrends of electrification, autonomous driving, and new business models such as ride-sharing, etc., have upended the industry. Recent Semiconductor shortages have made matters even worst. Some studies estimate a revenue loss of more than $200B this year alone due to the disruption within the Semiconductor sector. The automotive industry is responding to these disruptions in new ways that will profoundly impact their Intellectual Property Strategies. IP departments must find different ways to operate in these challenging times. The auto industry's operational dynamics are changing, and their corporate IP departments need a reset to optimize their IP strategy.
Disruption in the automotive industry
As outlined above, the global automotive industry is experiencing a massive disruption. The following are some key impacts of this disruption:
• Significant job cuts – Compared to conventional Internal Combustion Engines based automobiles, Electric Vehicles have fewer parts and do not require much maintenance. This means fewer jobs in making and servicing vehicles
• Diminished value of Intellectual Property – Auto companies, have invested substantially in R&D, generating large IP portfolios. However, electrification will make a large part of the portfolios technologically redundant and low commercial value
• New regulatory hurdles – Liability for accidents caused by autonomous driving is still a moving target and remains a significant source of uncertainty for auto companies
• Ever changing competitive landscape – Automotive industry generally is a well-oiled machine perfected by complex interactions between many suppliers and OEMs arranged in a tiered structure. However, new entrants such as Waymo, Cruise, Zoox, and Aurora are upending traditional OEM-supplier relationships. Further, consumer tech companies such as Apple are also rumored to be working on autonomous Electric Vehicles. This diverse mix of incumbents and new entrants coming from non-automotive fields is likely to complicate the competitive landscape
While the auto companies are grappling with these disruptions, the global shortage of Semiconductor devices has left them scrambling. Modern vehicles use a large number of Semiconductor components. Electrification, increasing adoption of Advanced Driver Assistance Systems (ADAS), and Smartphone infotainment systems are driving the Semiconductor demand in the auto sector. For example, management consulting firm Kearney estimates that Tesla Model 3 has 3.5 times electronic component density than the average internal combustion engine vehicle. Automotive is projected to be the fastest-growing market segment for Semiconductors.
This article will focus on this shortage, auto companies' response to the shortage, and the future of Intellectual Property development and management in the auto industry. Before we proceed, it is crucial to understand how the automotive semiconductor supply chain works.
How the Automotive Semiconductor supply chain works
The automotive industry follows a tiered structure (shown in the figure below) where OEMs integrate various systems or modules such as ADAS modules, Electronic Control Unit (ECU), or Telematic Control Unit (TCU) sourced from Tier 1 suppliers. Tier 1 suppliers, in turn, source components such as Microcontrollers, System on Chip (SoC) from Tier 2 suppliers. Tier 3 suppliers supply wafers and substrates and manufacture Semiconductor components such as Microchips for Tier 2 suppliers.
Reasons behind the global Semiconductor shortage (for the automotive industry)
One peculiarity of the automotive supply chain is the relatively large number of Tier 1 suppliers compared to a small number of Tier 2 suppliers. Designing Microchips is a technically involved task, and only a handful of Semiconductor companies can design Microchips that meet the complex technical requirements of modern vehicles. The number of Tier 3 suppliers, called foundries or fabs, is even lower. Although some Tier 2 companies can also manufacture Microchips, only a select group of foundries can build bleeding-edge Microchips. Furthermore, the Tier 3 suppliers do not make Microchips only for the automotive sector. These Tier 3 suppliers manufacture Microchips for Consumer Electronics, Telecommunications, Industrial, Aerospace, Computing, and Storage sectors, among others. Although the automotive industry is likely to witness the fastest growth in Semiconductor demand, it has to compete for Microchips.
With this background in mind, let us revisit the events of the past 20 months. As the world started feeling the impact of the Covid-19 global pandemic, automotive OEMs anticipated a steep decline in the demand for passenger vehicles. As millions lost their jobs and many others started working from home, it was reasonable to assume that fewer passenger vehicles would be sold. This forecast of reduced demand percolated down the supply chain, and Tier 1 suppliers adjusted their orders from Tier 2 suppliers. Tier 2 suppliers, in turn, reduced their orders from foundries. But recall that foundries manufacture Microchips not only for automotive companies. Therefore, they quickly reallocated their available capacity towards manufacturing Microchips for high-demand high-margin applications such as consumer electronics, networking, and data centers. The pandemic triggered massive Digital Transformation, the hallmark of which was remote working. It created enormous demand for networking equipment, laptops, video conferencing, and cloud-based collaboration services. With people spending more time at home, demand for entertainment services such as networked gaming and video streaming grew exponentially. This boosted the demand for content delivery and consumption devices such as Servers, Tablet computers, smartphones, etc. Foundries' reallocated capacity went towards producing Microchips catering to devices that enable remote working, communication, and entertainment. When the automotive demand came back, and it came back quicker than anticipated, foundry capacity was simply not available, and the lead times became more than 24 weeks. This created the global shortage about which we have heard so much.
The OEM Response
OEMs are responding to this "Semiconductor shock" very strategically. But before we dive deeper into their response, let us take a slight detour and briefly examine how the Semiconductor industry structure evolved from vertically to horizontally integrated.
The semiconductor industry was highly vertically integrated during the initial decades of its evolution. Companies such as IBM designed and manufactured their Microchips and further integrated them into systems sold to consumers. But in the late 1980s and 1990s, the pure-play foundry model came into existence and revolutionized the Semiconductor industry. It allowed so-called "fabless" companies to focus on niche and highly specialized areas without worrying about Microchip manufacturing and lowered the entry barriers. TSMC pioneered the pure-play foundry model and made vertical integration less lucrative.
However, very specialized use cases for Microchips are now driving a vertically integrated trend except for manufacturing (only a handful of companies can manufacture high-end microchips, which is unlikely to change). High growth areas such as Artificial Intelligence, Cloud, and 5G drive a tight integration between hardware and software to deliver optimum customer experience. Highly optimized AI algorithms running on bespoke accelerated hardware are becoming the norm. Similar vertical integration is likely to be seen in the automotive sector because mission-critical and complex use cases such as Deep Learning for real-time object detection in ADAS systems require close hardware-software interaction and optimization. This requires a tight and close collaboration between players operating at different tiers of the automotive supply chain.
Let us now return to the OEM response to "Semiconductor shock." OEMs are adopting various strategies such as building their in-house SoC design teams, partnering with foundries, licensing Semiconductor IP directly from IP providers, and partnering with Microchip design specialist companies. OEMs are also pursuing tight hardware-software integration by building their own Operating Systems, Security architecture, middleware, and Application Software optimized for the custom hardware platforms they will use. In other words, OEMs are exercising a lot more control over the hardware and software that go in their vehicles.
"People who are serious about software should make their hardware." – Alan Kay (American Computer Scientist)
Let's look at some recent initiatives from OEM on Semiconductors and hardware-software integration:
• In April 2021, VW's CEO Herbert Diess said the VW group plans to design and develop its high-powered chips for autonomous vehicles. Diess further spoke about achieving optimal performance through software-hardware integration.
• Tesla designs its custom chips optimized for its software and vice versa. Tesla developed a High-performance Computing (HPC) chip with Broadcom, produced by TSMC's 7nm process.
• Daimler and Nvidia signed a deal in 2020 to build a next-generation chip and software platform together. The developed software-defined architecture will be launched in 2024.
• On 18th November 2021, Ford said it signed a non-binding agreement with GlobalFoundries for joint R&D and manufacturing of chips to supply its assembly lines and others in the US automotive industry. Ford's CEO Jim Farley wants Ford to become the world's largest electric car maker, and Semiconductors are critical to his aspirations.
• On the day Ford announced its partnership with GlobalFoundries, General Motors announced it was "working with seven chip suppliers on three new families of microcontrollers that will reduce the number of unique chips by 95% on future vehicles." Qualcomm, STMicroelectronics, TSMC, Renesas, NXP, Infineon, and ON Semi are supplier partners.
Implications for the IP Strategy
Auto companies' changing business strategies necessitates a corresponding change in their IP strategy. Here are a few recommendations:
1. Better management of IP risk:
Currently, OEMs are not exposed to significant IP risk, except Non-Practicing Entities (NPE). The auto industry doesn't see many IP disputes among its participants. However, as OEMs design their Microchips, they will be exposed to increased infringement risks. Complex patent litigations and licensing are commonplace in the Semiconductor world. Tighter hardware-software integration will also bring in additional risk as core patents on computer architecture, hardware optimization/acceleration, operating systems, middleware, etc., are held by non-automotive companies.
2. Developing IP with monetization in mind
OEMs traditionally focused only on patents that covered the vehicle functionalities. These patents had little to no value outside the auto industry. However, if OEMs develop Semiconductor, Software, and System patents, then such patents are likely to be widely applicable in other sectors. This dramatically opens up the monetization avenues. OEMs should ensure the new patents covering Semiconductors, Software, and Systems are drafted in a way that enables their applicability outside the auto industry.
3. Tier 1 suppliers will need to enforce their IP:
Currently, OEMs are not that involved in negotiating with Semiconductor companies in Tiers 2 and 3. This will change. OEMs are now partnering with Semiconductor companies, thereby taking a large part of Tier 1 suppliers' role. Similarly, non-automotive companies are playing in the Tier 1 territory. Since Tier 1 suppliers stand to lose substantial revenue, they are gearing up to reposition themselves as a solution provider and not just a module provider. On the IP side, Tier 1 suppliers need to ensure new partnerships between OEMs and Semiconductor companies respect their IP. Tier 1 suppliers should also provide non-auto players in Tier 1 territory pay their fair share of royalties to Tier 1 suppliers. In a nutshell, Tier 1 suppliers need to shift their mindset when enforcing their IP and to capture total value from their IP.
4. Time to upskill the IP department
IP departments need to upskill. Current IP departments in auto companies are heavily focused on mechanical, electrical, and materials science knowledge. However, they need to acquire skills in Semiconductor and Software technologies. Semiconductor licensing programs are very different from auto licensing programs. Identifying infringement is the biggest challenge in Semiconductor licensing, and corporate IP departments need the right partners who are experts in Semiconductor Reverse Engineering. IP departments in auto companies need to invest in strategic partnerships with companies with IP and Reverse Engineering expertise.
5. Applying a different patent drafting strategy
As noted above, detection of infringement of Semiconductor patents is generally tricky and requires substantial reverse engineering. Patent claims should be drafted, keeping infringement detection in mind. Specifically, patent claims should be written such that the infringement can be shown by careful reverse engineering. It is recommended to partner with reverse-engineering experts to thoroughly understand what patent claims are easiest to enforce.
Anand Rohit works as a Director of IP Solutions group of UnitedLex. ULX works with the world's top automotive and Semiconductor companies as their strategic partner. ULX offers an unparalleled combination of deep technical and subject matter expertise. ULX IP expertise and reverse engineering capabilities allow clients to defend their business practices.