Ducati Corse, a subsidiary of Ducati Motor Holding S.p.A, was responsible for managing the activities of the Ducati racing teams. Historically, Ducati had been the dominant force in the Superbike Circuit winning 13 titles since 1990. Until recently, Ducati was only part of the World Superbike series. But a recent rule by FIM (in 2001) permitting the use of 4-stroke engine in bikes as against 2-stroke engines, which was a mandate earlier (for MotoGP) prompted Ducati to get into MotoGP. Despite the fact that high costs were involved and Ducati faced the risk of being a late entrant into MotoGP, the Ducati team was highly enthusiastic about this new development. Ducati had a fairly successful first season (2003) but the 2004 campaign turned out to be disappointing. Determined to come back strongly, Filippo Preziosi, (Technical Director) and the team decided to move from an integrated design approach to a modular design for its next generation bike (GP5).

Why did Ducati Corse succeed in its first year of MotoGP competition?
Even though Ducati entered the MotoGP arena in 2003 as underdogs with very limited experience, they literally made their presence felt. They went on to achieve 14 front-row starts, 9 podium finishes, 3 pole positions, and even tasted one victory. One of the key reasons that I believe led to their success was the inner zeal, commitment and passion that the Ducati engineers displayed in designing the first GP bike. Few of the other important reasons for their success included: 1. The ‘Ducati Method’: The well structured Ducati Method was one of the primary reasons for the success of Ducati. It laid emphasis on data-driven approach to problem solving, data analysis through computer simulations, feedback from racers, clear demarcation, prioritization and ownership of problems, face-to-face communication, open office layout , weekly design reviews etc. 2. Experience: Years of prior experience in terms of developing commercial bikes and bikes for Superbike series gave them the confidence to build the GP3 bike. 3. Partnerships and Relationships: Close relationships with top racing companies as Ferrari, Toyota etc. and access to an array of specialized bike parts suppliers also contributed to the success of Ducati. 4. Smart Decisions: Even though the decision to get into MotoGP was made as early as 2001, Ducati’s management team decided to participate in the race only in 2003 and in the process bought themselves some time to come up with a solid and competitive bike. 5. Top recruits: Ducati also benefited because of the great resources that they had. This included their ability to attract top engineering graduates in Italy, presence of two of the top racers in their ranks - Loris Capirossi and Troy Bayliss. 6. Sound leadership: A significant contribution to Ducati’s success must be attributed to the able and thoughtful leadership of Domenicali, Preziosi , Cecchinelli. 7. Passionate employees: Ducati Corse consisted of some very passionate and dedicated engineers who were all crazy about building superior bikes and engines. 8. Low expectations: The fact that they were competing in MotoGP for the first time, Ducati went ahead to the championship with very low expectations. Perhaps this eased the pressure off the racers which enabled them to go all-out with nothing to lose. 9. Culture: Presence of a relatively informal culture wherein the employees rode motorcycles to work and dressed casually made the employees always feel at home. 10. Others: Some of the other important factors included presence of cross-functional teams for specific issues, regular meetings to discuss the progress of the project, adequate use of technology etc. Also huge importance given to physical testing which involved dyno bench tests, wind tunnel tests etc. and track testing across 10 different circuits around the world ensured that the resultant GP3 was perfectly ready for the MotoGP race.

What went wrong/different in the second season?
Riding high on the success of 2003 season, the Ducati team wanted to take their success a step further. In their efforts towards achieving the optimum design, they ran several tests to identify any design related issues and even hired 2 engineers to do the data analysis. Through these initiatives and with the inputs from rider feedback and simulation tests, Ducati identified some weight distribution related issues with GP3. As a result, when finally GP4 (devoid of these issues) debuted, expectations were high. But, unfortunately the 2004 results were not satisfactory enough. Few of the things which might have gone wrong include: 1. Wrong decisions: The decision by Ducati to wait for 2003 data to arrive seems to have backfired as this resulted in less time being available for the development of GP4. 2. Data issues: As the 2003 season progressed, the heap of data began to grow multi-fold and it became simply impossible for Ducati to analyse such huge dataset despite hiring 2 additional engineers. Moreover, in certain cases some of the data was contradictory, for eg. the rider’s feedback at times did not match with the data obtained. 3. Over-confidence: The unexpected success in 2003 made the Ducati team over-confident which might have hurt them. In the words of Preziosi “We made a big mistake. We changed a lot to try to solve problems. We were a bit over-confident”. 4. High expectations: Because of high expectations from riders after 2003 success, the pressure to perform better in 2004 might have taken a toll on them. 5. Parallel projects: The fact that parallel projects, which included improvements in GP3 design for the next GP4 version and the development of GP5 for 2005 season (which had also started) had put additional stress onto the engineers. Further, poor weather conditions delayed the winter tests which had put further pressure on the engineers. 6. Change in design strategy: The disappointing performance also led Ducati to move away from its normal course and bring about design changes midway during the season which Preziosi likened to “changing the wheel of the train while it is moving”. 7. Major design changes: Also, in the first place, Ducati possibly did not need to bring in so many changes (60% new components in GP4 Vs GP3). In Cecchinelli’s words “….You do not look at the data because you want to understand why you are performing well”.

How did Ducati Corse use data on performance to learn?
Ducati essentially used the ‘Ducati Method’ for the bike development which encompassed the following 3 principles- a) data-driven approach to problem solving, b) clear priority in solving problems and c) frequent face-to-face communication
Data-Driven approach to problem solving - Ducati effectively used data related to lap times, top speeds, average speeds etc. to assess how its bikes were performing vis à vis its competitors. Additionally the on-board sensors provided data on engine speed, oil, engine and tire temperatures etc. The engineers also received feedback from riders themselves. All these sources of data were used in identifying any mechanical problems or design related issues for a particular subsystem. Apart from these, the engineers generated their own data through extensive simulations which was also used to run virtual experiments.

What is your assessment of their approach to learning?
The ‘Ducati Method’ provides a very good platform for learning. As explained above, the data driven approach enabled the identification of the problems. These problems were then prioritized, at a high level by Preziosi himself and at a lower level by the chiefs of various departments. Weekly design review meetings, cross-functional teams, open office lay-out and emphasis on face-to-face communication further enabled the process of addressal of these problems. In a way, this approach suggests that changes in one area positively affect another area and so on. The cyclical learning process essentially starts at collection of data and rider’s feedback -> identification of issues -> prioritization of issues -> design changes -> change in race results -> collection of data and rider’s feedback and so on. It can be best represented pictorially as per Figure 1 (Appendix)