lA a Formé 12 000 Ingénieurs dans le Monde… Sa Vision sur les Marchés va vous Surprendre!

Daniel Crob, a Normal Sup graduate with a background in mathematics and computer science, has dedicated over two decades to specializing in complex systems. His career began in academia, where he focused on theoretical computer science, discrete mathematics, and mobile telecoms. A pivotal experience was working for seven years with Jean-Louis Chetter, the "father of GSM," who had standardized the radio layer protocols for mobile communication. Chetter, a hands-on engineer who once coded 95% of a new mobile solution himself, eventually left Nortel Networks after criticizing its strategy. This experience provided Crob with a deep understanding of mobile radio networks. After his academic career, Crob ventured into entrepreneurship, founding a firm called Sésame over 15 years ago. This firm, now employing around 40 system architects, specializes in system architecture, a discipline Crob essentially introduced to France. System architecture involves designing the overall balance within a complex system to ensure its detailed components function harmoniously. For example, in a car, this means ensuring that the six main subsystems (engine, body, software, etc.) are compatible and work well together, overcoming the inherent "siloing" within industries. The goal is to prevent issues like a fuel pump problem affecting the car's windows due to software interdependencies. This discipline originated in the United States in the 1960s, driven by the logistical challenges of World War II and the need for planetary-scale command and control systems during the Cold War. The methods developed by the Americans for engineering large-scale systems were later civilianized by NASA, leading to the "NASA System Engineering Handbook" in the 1970s. This knowledge became widespread in Anglo-Saxon universities, but in Europe, particularly France, it remained largely unknown until the early 2000s. Crob had the opportunity to lead an industrial chair in complex system engineering at École Polytechnique in 2002, funded by Thales. This allowed him to develop methods and build relationships within the French industrial ecosystem. Realizing the limitations of the academic system, he decided to launch his own company around the age of 50. His firm, Sésame, quickly became a leader, securing a major training program for Safran engineers. Crob recently sold Sésame, not due to lack of success, but because he thrives in creation and finds routine management less engaging. He sees himself as a good manager during creation periods but less so when operations become standardized. Regarding French industry, Crob offers a nuanced perspective. While some sectors, like aeronautics, remain global leaders (Airbus, Dassault, Safran), others face challenges. Safran, for instance, is a world leader in aircraft engines and components, manufacturing almost everything for an aircraft except the fuselage. Developing aircraft engines is incredibly complex, with only a handful of companies worldwide possessing the expertise. Even Safran, a leader, can encounter difficulties, as seen with their attempt to develop a new private jet engine. The conversation also touched on the rising Chinese competition, particularly in semiconductors and aerospace. While China is investing heavily in these areas, their approach often involves reverse engineering. For example, Chinese company Comac's C929 and CR929 are essentially copies of the A320. However, merely replicating an aircraft's physical components doesn't guarantee success. An aircraft requires a certificate of conformity, which involves a rigorous certification process demonstrating safety (e.g., a catastrophic event probability of 10^-7 per flight hour). This process generates massive documentation, proving the system's reliability, which the Chinese struggle with because they don't fully understand the underlying safety design choices. Airbus's design choices, such as quadrupling data transmission chains, are driven by safety requirements, leading to complex systems. Crob's work as a system architect involves making crucial trade-offs between cost and safety. He emphasizes "doing right the right system," meaning not only building a system correctly but also ensuring it meets a genuine need. He shared anecdotes about the French Ministry of Defense specifying vehicles be painted green, even for desert operations, highlighting how the original "why" can be lost over time. Contradictory requirements, often politically driven, can also lead to systems that are impossible to build or become prohibitively expensive, as seen with the A400M military transport aircraft. Poor architecture can also make systems resistant to evolution; for instance, attempts to make an American fighter jet more robust by replacing aluminum with titanium led to unforeseen balance issues and widespread system changes. Having sold his architecture firm, Crob is now embarking on a new venture in industrial simulation. He explains the different stages of industry: innovation, engineering design (where France excels), manufacturing, and industrial operations (maintenance, spare parts). He highlights that France remains competitive in design and engineering across the entire chain in sectors like aviation and nuclear. However, the automotive industry faces disruption, primarily due to Tesla's influence. Historically, cars were mechanical, with electronics gradually increasing. Tesla, however, re-envisioned the car as "software on wheels," a smartphone on wheels. This shift moved the industry's center of gravity from mechanical engineers (motorists) to software engineers. European manufacturers, with their entrenched mechanical paradigms and fragmented software development, struggled to adapt quickly. Chinese manufacturers, starting with a clean slate, embraced this new software-centric paradigm from the outset, designing simpler, more efficient core software architectures and reducing cabling complexity. Crob's training methods for system architects are highly practical, focusing on "learning by doing." He trains experienced engineers, not fresh graduates, because system architecture requires a deep understanding of industrial design and the ability to make impactful decisions. His work involves coaching clients through complex problems, even when he himself is initially unsure of the solution, maintaining confidence while figuring things out. He also notes a competitive advantage for Europe in certain complex industrial systems compared to China. While the Chinese are pragmatic and learn quickly, their bottom-up approach to design has limitations when it comes to certifying complex systems like aircraft. They can build an airplane, but they cannot yet certify it or produce 60 per month, indicating a gap in understanding the underlying safety and industrialization processes. Crob highlights the importance of identifying a limited number of key parameters in complex systems. In aircraft design, for example, two or three key parameters like range and passenger count dictate millions of other design choices. He mentions the "777 rule" in ergonomics, suggesting that the human brain can immediately comprehend a maximum of around 350 (7x7x7) organized objects. This implies that there are typically around 7 key parameters for understanding and managing complex systems. He observes a recurring pattern of "7" in various contexts, from historical cycles to the number of subsystems in "world models," speculating it might be linked to intrinsic limitations of the human brain, much like human dimensions define architectural spaces. He emphasizes the excellence of French engineers, often underestimated, especially when compared to German manufacturing or financial skills. French industrial companies, even those with limited production facilities in France, maintain a global presence and critical mass, like Schneider Electric, for whom Crob has trained diverse international teams. He notes that American trainees were initially the most challenging, often taking a condescending attitude towards a French trainer, until they experienced the practical benefits of his methods. Crob concludes by noting the fundamental difference between engineered systems and financial systems. Engineered systems strive to reduce complexity by decoupling subsystems, while financial systems are characterized by intense, interconnected feedback loops and high noise levels, making it difficult to isolate useful signals. Despite the challenges, Crob finds it encouraging that France possesses strong industrial talent, capable of innovation and entrepreneurship, even when navigating complex relationships with the state.