The future manufacturing costs of emerging technologies are difficult to assess because of the complex dynamics of both product and process innovation, and because cost data often are proprietary and difficult to obtain. One method of forecasting potential future technology costs uses the concept of manufacturing progress functions, which are closely related to manufacturing experience curves. Manufacturing cost is related to cumulative production volume for a specific firm in an industry, using relatively simple relationships that have been observed for a wide range of products. Progress functions and experience curves take into account scale economies, technological improvements in production processes, improvements in product design, and improved efficiency of workers and production management. Here we analyze the future manufacturing cost of an innovative new technology – the automotive PEM fuel cell system – using a manufacturing progress function analysis. Based on manufacturing progress function theory and the assumptions used in the analysis, we trace the manufacturing costs of fuel cells systems from present levels of $1,500 – 2,500 per kW to less than $50 per kW. For products such as PEM fuel cells that may reach high levels of accumulated production, we suggest methods for bounding forecasts in order to guard against eventually forecasting unrealistically low costs.

The future costs of electric-drive vehicles, like those of any new technology, are uncertain. One method for forecasting cost reductions uses the concept of the ‘experience’ curve. Experience curves take into account sale economies, technological improvements in production processes, improvements in product design, and improved efficiency of workers and production management. Here we analyze the future manufacturing cost of a brushless permanent magnet electric vehicle drivetrain using experience curves and a Monte Carlo simulation technique. Our preliminary forecast is a drop in manufacturing cost from today’s $13,000 (with about 2,000 units of cumulative production) to a cost of about $1200 when full scale economies and ‘learning’ have been realized. In an ongoing study at the University of California, Davis, experience curves are being integrated with a detailed vehicle cost model to develop more sophisticated cost forecasts for complete electric vehicles.