It is clear that artificial intelligence is disrupting every industry as we know it. This includes not only the sectors that have garnered the most attention — such as SaaS, fintech, healthtech, and travel — but also traditionally heavy industries that are ripe for disruption.
As an industrial AI-oriented investor, I have witnessed how many companies in the field are increasingly embracing automation and data-driven decision-making, and how their approach may vary based on both what the corporation needs and the resources they have available.
In this piece, I will discuss various options firms have to integrate AI-powered technologies into their business processes, and highlight both the pros and cons I’ve observed in each of them.
1. Establish an internal R&D department
A route several companies take is establishing their own R&D department to develop AI technologies. For instance, Siemens, through its AI Lab, is pioneering various potential applications of industrial AI.
While Siemens has been able to reach some breakthroughs — such as reducing production times without the need for new hardware — the reality is that for most firms, the benefits they can derive from an internal department are limited.
Unlike startups, the corporate world has slow processing times, low tolerance for errors, and high expectations that can kill projects before they harness their full potential. Startups, on the other hand, are adept at pivoting and know that several iterations are required before finding a real breakthrough, especially with technologies like AI that require us to be in a constant “learning” mode.
This is why, from my perspective, companies that opt to leverage this approach need to give that department autonomy so that it can operate like a startup. Otherwise, the sluggish pace with which corporations traditionally operate will likely hinder their prospects.
2. Create a corporate venture fund (CVF) or accelerator that focuses on AI
Behemoths like Toyota — initially through the Toyota Research Institute, and then through Toyota Ventures — and Qualcomm, through Qualcomm Ventures, have poured hundreds of millions of dollars each by investing in promising startups in AI, robotics, and other frontier technologies.
On the other hand, other firms — like Fujitsu, through the Fujitsu Engineering Accelerator, or Volkswagen, which partnered with well-known Silicon Valley accelerator Plug and Play — have created proprietary acceleration programs to support emerging ventures that focus on the needs and challenges of their industry. There are benefits to this, since they can help firms pilot projects with startups and leverage their resources to help these startups succeed.
Nevertheless, this approach has limitations too. Establishing a venture fund or accelerator does not change a corporation’s deeply-ingrained culture. Furthermore, the operation of these funds is usually constrained by additional factors, such as protocols and rules established by the parent company. Traditional corporate processes can also clash with what is needed to develop breakthrough AI technologies.
3. Hire a Chief Digital Officer (CDO)
This step involves hiring an individual or forming a department that will be tasked with digitizing the company. These responsibilities will encompass developing AI adoption strategies and liaising with startups. The Chief Digital Officer (CDO) will also focus on enhancing efficiency, competitiveness, and growth through digitization.
Potential drawbacks of this in-house approach relate to the fact that startups might find it challenging to communicate with corporate employees, because they are accustomed to different business models and have completely divergent communication protocols. Additionally, the CDO might rely on their existing network of contacts for potential partnerships, limiting the scope of effective collaborations.
Another consideration is that the CDO needs to be aligned with the company’s overarching vision. For instance, if the CDO wants to drive fast transformation, and the firm is not ready to progress at that pace, projects might stall, and only lead to further frustration.
In general, this model works better when the corporation interacts with a VC fund, since a venture capitalist can quickly understand which of their portfolio companies is better suited to solve a particular need or problem.
4. Organize AI-themed hackathons
Recurrent hackathons — for example, annually — are a powerful method to generate new ideas and solutions. Nowadays, this strategy is not only implemented by corporations, but also by startups and funds. I have personally used this approach, and one of my portfolio companies regularly organizes hackathons, since they provide an extraordinary platform for people to be creative and think outside the box.
Historically, some products created at hackathons have gone on to become great successes. For example, at one event organized by Schneider Electric, participants developed an AI-powered solution to optimize energy management systems. Schneider Electric took this prototype and further developed it, benefiting from more efficient energy usage and eventually passing on these cost reductions to its customers.
In the same vein, a GE-hosted hackathon spurred the development of an AI application that improves wind turbine efficiency by analyzing operational data and automatically adjusting control settings. GE expanded on this technology, and now, it optimizes the wind farm operations of GE’s renewable energy division. It is one of many solutions developed at hackathons that GE has eventually implemented.
Bosch’s “Connected Experience” hackathon, which focuses on AI and IoT innovations, is another great example of an AI-centered event by an industrial company, and it is expected that the creations that emanate from it will accelerate disruption at the firm’s manufacturing and automotive divisions.
The secret to a successful hackathon lies not only in the ability to organize it and the willingness to invest time and money but, more importantly, in understanding why you are doing it and how to utilize the results—the ideas generated by the participants. On one hand, it’s crucial to allow participants the freedom to think creatively, as the essence of a hackathon is in the search for new ideas. On the other hand, systematizing the results is necessary. Mastering this balance can make a hackathon an excellent source of new technologies for the company, or talent, because a hackathon is not only a platform for discovering new technologies but also for identifying individuals capable of developing these technologies within the company.
Final thoughts
While these four approaches can be potentially successful strategies for corporations to integrate AI technologies into their processes and improve results, I must remark that a common thread here is the importance of communication and understanding between two radically different ways of working.
AI startups and innovators can often find it challenging to communicate with corporate employees, therefore, this is a skill that needs to be taught, since effective communication can pave the way to success.
Hence, a final recommendation for a corporation is to have an employee at the company that can work with startups and teach them how to bridge this communication gap. Google is a positive example of this. I met someone at Google who, besides being involved in enterprise sales, was a mediator who taught startups to find common ground with large conglomerates. This is key, since reshaping today’s industries with the power of AI will require us to work together despite our differences, and those who do not know how to collaborate will likely be left behind.
