How well do experience curves predict technological progress? A method for making distributional forecasts

Lafond, François; Bailey, Aimee Gotway; Bakker, Jan David; Rebois, Dylan; Zadourian, Rubina; McSharry, Patrick E.; Farmer, J. Doyne

doi:10.1016/j.techfore.2017.11.001

Cited by 52 publications

(50 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similar to [1], let us consider that empirically cumulative production growth follows a smooth exponential behaviour in the presence of noise, by assuming that production is a geometric random walk with drift g and variance σ 2 a . Within this model, cumulative production arXiv:1708.02605v1 [q-fin.GN] 1 Aug 2017 2 is given by:…”

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

“…variables, with mean zero and variance σ 2 a . For the calculation of cumulative production and its volatility in [1] the saddle point method was used. The main idea of the saddle point is to approximate an integral by taking into account only the range of the integration where the integrand takes its maximum.…”

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

“…A priory, this can only be correct for small variance σ 2 a . In [1] first the expectation value of cumulative production and its variance were calculated, which lead to the multiple integral over a i…”

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

“…The aforementioned work had a major impact on the development of this concept, by arguing that technical learning was a result of ex-perience gained, based on the idea learning by doing. Some researchers question its usefulness for forecasting and planning the deployment of industrial and technological activities [1,[11][12][13][14]. In the aforementioned literature it has been found that experience curves can be used to estimate future technology costs, considering the shape of the forecast error distribution.…”

Section: Introductionmentioning

confidence: 99%

“…

In this paper we study the volatility and its probability distribution function for the cumulative production based on the experience curve hypothesis. This work presents a generalization of the study of volatility in [1], which addressed the effects of normally distributed noise in the production process. Due to its wide applicability in industrial and technological activities we present here the mathematical foundation for an arbitrary distribution function of the process, which we expect will pave the future research on production and market strategy.

…”

mentioning

confidence: 99%

See 4 more Smart Citations

Exact probability distribution function for the volatility of cumulative production

Zadourian

Klümper

2018

Physica A: Statistical Mechanics and its Applications

Self Cite

View full text Add to dashboard Cite

show abstract

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

Section: Volatility For Narrow Distributionsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…

…”

mentioning

confidence: 99%

See 3 more Smart Citations

Exact probability distribution function for the volatility of cumulative production

Zadourian

Klümper

2018

Physica A: Statistical Mechanics and its Applications

Self Cite

View full text Add to dashboard Cite

show abstract

Technological learning: Lessons learned on energy technologies

Haas

Sayer

Ajanović

et al. 2022

WIREs Energy & Environment

View full text Add to dashboard Cite

The concept of technological learning is a method to anticipate the future development of the costs of technologies. It has been discussed since the 1930s as a tool for determining manufacturing cost reductions, starting in an airplane manufacturing plant, by means of learning curves and has been widely used since the 2000s in energy models to endogenize technological change. In this paper, the theoretical concept of technological learning based on energy technologies is analyzed based on examples from the literature. The main low‐carbon power generation technologies, photovoltaics, concentrated solar power, wind and nuclear energy were analyzed, showing different cost trends. Additionally, the impact of policy support on technological learning was discussed in concrete examples of bioethanol and heat pumps. We find that the homogeneity and the modularity of a technology are essential for high learning rates. A good proof is the manufacturing cost development of photovoltaics in recent decades, where a rather stable learning rate of 20% has been identified. On the contrary, nuclear power did not evolve into a homogeneous technology due to required environmental adaptations caused by accidents and the lack of standardization and application of new engineering approaches. In that case, the overall price further increased. Finally, another important condition is stable legal and regulatory conditions regarding the implementation. This article is categorized under: Policy and Economics > Green Economics and Financing

show abstract

Spacecraft for Deep Space Exploration: Combining Time and Budget to Model the Trend in Lifespan

Tsai

Berleant

Segall

et al. 2021

Proceedings of the Future Technologies Conference (FTC) 2021, Volume 3

View full text Add to dashboard Cite

How well do experience curves predict technological progress? A method for making distributional forecasts

Cited by 52 publications

References 57 publications

Exact probability distribution function for the volatility of cumulative production

Exact probability distribution function for the volatility of cumulative production

Technological learning: Lessons learned on energy technologies

Spacecraft for Deep Space Exploration: Combining Time and Budget to Model the Trend in Lifespan

Contact Info

Product

Resources

About