2025 - a transition year in the automotive and software industry

We are close to year-end, and the election season's major event is happening next week: the elections of the US president, the Senate, and a large part of the Congress. This has a major impact on every aspect of global politics and economy.

I am summarizing for my readers some assumptions about the changes that might happen next year in the automotive and software industry. These are mostly speculations and are consolidated based on my reading and the podcasts I listen to. However, I will not list here the references in much detail, I will point to some people I follow regularly.

The major influence of the election outcome next week will be related to technology investments. There is a consolidated opinion that Donald Trump represents the interests of several so-called techno-kings and the policies that his administration would promote will be more in favor of such technology companies. Elon Musk is the most prominent exponent of the technology conglomerate and his companies recently showed major technological advancements in space, robotics, and cloud technologies. Still, his approach and announced involvement in the government would definitely support the announced measures for deportation, slashing major chapters of investment and increasing tariffs for the entire world. It will be a major difference in the application of the corporate tax, with Donald Trump supporting a slash of the 21% charged today and Kamala Harris looking to increase it to 28% and support the small entrepreneurs.

The impact of mass deportation will have an impact on the attractiveness of the United States for highly skilled migrants and how many exceptional people will do their research in the United States. Two major countries for technology research are not in good economic or political terms with the United States: China and Russia. India and Latin America are the major suppliers of migrants for the United States, which still needs an extra workforce to support the economy.

The second significant impact will come from the tariffs. The United States is the only major economy that imports more than exports. This is largely due to the attractiveness of the financial systems and the interest of the world to buy American assets. The American economy imports products and the capital market exports assets. The Chinese economy has a supra-production issue, largely due to their mode of economic growth through investment in production capacities, but owns a large part of US debt. The tariffs would generate inflation, as it is very less likely that the taxes will be absorbed by the distributors in the US, will put pressure on the major European brands, and would potentially affect the European automotive industry, but much less the Chinese brands. VW, Mercedes, BMW, and Volvo have major production facilities in the US and might dodge the bullet, while the Chinese will definitely build plants in Mexico.

The investments in technology will continue, regardless of the name of the new president, and will need to present some monetization of the huge investments made in the AI trend. OpenAI spends more than it can collect from the 100+ million users of ChatGPT and the companies that are building on top of the LLMs. The advancement of OpenAI competitors (Google Gemini, Anthropic Claude) or solutions like Perplexity.AI will definitely generate a lot of options to build applications that could monetize AI technology. There are plenty of things that are coming on top of open-source LLMs like LLAMA or Mistral and the trend is just beginning. Nate Jones has very interesting opinions about what value we can get from AI-powered search compared to the traditional one, namely that the way information is displayed matters and still ChatGPT, Gemini or PerplexityAI product produce a lot of clutter. Sometimes the good old UX is more helpful, like when you are looking for information related to the movies running in theatres or football games tonight.

The other dimension influencing investments in technology is the way states will support technology. There is a huge interest in military technology, where the high tech needs to be doubled by extensive production capacities. There will be the need to invest in the diversification of the power sources and balancing of the grid plus in the update of the power grids in the areas where solar panels are mounted in the residential areas. Robotic technology will definitely be improved. All these will influence software development in a way that was not seen before: the companies which are having these products MUST expand with bigger or smaller software R&D departments, able to design and maintain the products. A big part of the software business will be the maintenance of the solutions in the field. That is part of the software-defined products that need feature updates, security updates or bug fixing.

Software-defined products are already in the market, but you do not see it so obviously. The NEST thermostat developed by Google is primarily a solution software-driven, as thermostats have already been available in large supply with wireless connectivity since 2010. Amazon-developed solutions for home surveillance get a huge set of features from AI and promise to strike the right balance with privacy. Several product types will probably be enhanced with the help of trained models that can run on edge devices, based on the data that can be collected with the wide implementation of 5G technology. Definitely, ,most of the products will require full stack implementations, with an interesting mixture of back-end services, most likely connected to some sort of AI model, connectivity over 5G on the device, AI models running on edge devices and enabling the actuation of hardware or reading of the sensors mounted on the product. Of course, the constant update of the software on the edge devices will be a must.

The trend will be most likely implemented by the automotive industry, which will take advantage of the transition to large ecosystems on the OEM side, where large data lakes will be used to harness features for end customers. One of the most difficult challenges is to transition the ecosystems of automotive producers toward an architecture that would enable the software features as the primary driver of sales. This is possible and was proven by the new solutions implemented by Chinese OEMs. Huawei AITO and Xiaomi SU7 are examples of solutions driven by challengers of the traditional producers of the cars. The most hailed features are around connectivity and the seamless transition from car to phone to home appliance devices. Plus many many screens and ADAS features. The advantage of Huawei and Xiaomi comes from their huge ecosystem of devices and their experience in designing experiences for a market (China) that is younger and more digital-oriented than Europe or the US. We should not overlook the initiative from Google to provide Android AUTOMOTIVE, with integrated Google services. It remains to be seen the impact of the age of buyers in the adoption of the full digital features in the Chinese cars, as the European adn American drivers could be more conservative than expected.

Connectivity to the street infrastructure will be very interesting support for the cars, one can expect that this will be more available in the 6-700 major cities and less available in remote locations. Linked with the connections to the backend servers of the OEMS, this limitation will drive the separation of experiences in different geopolitical spaces and most likely will generate an improved experience when the edge devices have good solutions or trained models.

The major trend in the development is that OEMs are reclaiming the space of software development and creating large organizations. BYD employs 105.000 engineers in their RD departments, Chery, BAIC Motor and Anhui Jianghuai Automobile Group partner with Huawei, Volkswagen's CARIAD employs 6500 software engineers and teams up with Rivian, BMW employes close to 10.000 software and IT engineers and scales in partnership with Tata Technologies and NTT Data, Stellantis aims for 4500 engineers for their software business and the examples will continue. Toyota is looking for 18000 engineers, but also intends to harness the cooperation with DENSO and Woven, with a dedicated Software Development Center.

These moves will make more difficult the life of the traditional Tier 1 suppliers such as Bosch, Conti, Aptiv, Magna, Forvia, Valeo and Visteon. These companies will be excellently positioned to further develop hardware, but the trend of separation of software from hardware will trigger further changes in their organizations. Bosch's ETAS and Conti's Elektrobit see already a shift in their connection with mother organizations. The zonal architectures are not yet implemented, the demand for software functions is reduced compared to the size of the organizations and the monetization models are not sufficiently mature. After the significant push coming both from OEMs and microprocessors manufacturers in 2020-2023, we see now another hurdle for Tier1s, when the traditional areas of execution are split between OEM and other new players in the market. Most of these Tier1 companies are going through a major redefinition of the workforce, which includes layoffs and refocusing of their existing business units.

However, one of the significant models of development will remain the co-creation of software solutions and more adoption of open source in the cars of the future. This should also include back-end solutions, non-differentiating software, and some open architecture elements to enable other players to plug their digital products into the car.

The geopolitical impact of the elections will be noticed in software ecosystems, where several attempts of the United States to regulate the access of software in the country would harm the Chinese OEMs to enter this big market, a thing that could be replicated in reverse by China or Europe to protect local producers.

One should not forget that overall there is an overcapacity for production of the cars, which could play in different ways. Some Chinese OEMs could take over or build capacities in Eastern Europe and Mexico to circumvent the tariffs, while some European or American OEMs (e.g. Stellantis or Mercedes) will partner with Chinese OEMs to catch up with the delay in the electrification technologies and production strategies. Geely just launched a PHEV model (NordThor 2.0) with more than 2000km of autonomy and this is a very high bar to compete with.

I see several important things happening in 2025:

• OEMs will advance in their strategies to increase sales and focus on electrification and SDV, but most likely they will not commit to the hard constraints to reduce the product of traditional combustion engines to zero by 2030
• There will be a war on tariffs and production sites that would put prices a bit down after the increase happening in 2021-2023. This will positively affect the consumer but will delay the investment in RD. I hope that I am wrong with the second part.
• The different attempts of the Western OEMs to catch up with the digital solutions will be accelerated, as they are directly competing with Chinese in 2 fields (electrification and SDV)
• The layoffs will be finalized for the German companies and will affect in addition the companies which deliver engineering services. A new baseline could be expected in this field, with the major competitors from India pushing harder in a market dominated by the need of decreasing the prices. e companies providing services
• Chinese companies will setup operations or co-operations with European Engineering Service Providers to penetrate local market for software and technology.

I am looking forward for your comments and opinions. Get in touch with me if you need a helping hand in the transition or just to exchange opinions.

Read more from my documentation

Stellantis to Spend $34 Billion on Software and New Vehicles - The New York Times

Stellantis Targets $20B+ in Sales From Software by 2030 | IndustryWeek

Software-Defined Vehicles (SDVs): The Road Ahead for Asian OEMs

The BMW Group and CRITICAL Software are expanding their cooperation

BMW Group and NTT DATA Romania sign Joint Venture contract: Development and operation of business IT solutions in focus.

Toyota's Key in Software Survival: In-House Capability and Real Customer Contact

TMC Announces Changes to Organizational Structure, Executive Structure, and Senior Professional/Senior Management Employees | Corporate | Global Newsroom | Toyota Motor Corporation Official Global Website

Geely overtakes BYD with the EM-i NordThor plug-in hybrid system

People I follow to have insights :

Dr. Moritz Neukirchner , Augustin Friedel ?? ?? ?? ?? , Florian Rohde , AUTOSAR , Dr. Fatih Tekin