A Cognitive Model for Technology Adoption

Fariborz Sobhanmanesh, Amin Beheshti, Nicholas Nouri, Natalia Monje Chapparo, Sandya Raj, Richard A. GeorgeAlgorithms 2023 · DOI · OpenReview

TL;DR

This paper introduces a novel data-driven pipeline and a parametric mathematical model (ETASC) to predict the adoption rate of emerging technologies at country and industry levels. By integrating global indicators and a cognitive approach, it aims to provide policymakers and businesses with better tools for future planning and decision-making in the face of rapid technological change.

What Problem Does This Paper Address?

The rapid advancement of technologies like AI, Machine Learning, and Robotics, often termed Industry 4.0, is transforming work and society. However, there's a significant gap in our ability to accurately forecast the adoption rates and subsequent impacts of these technologies. Existing models often focus on individual technologies or lack a comprehensive, global perspective that accounts for diverse national and industrial contexts. This makes it difficult for policymakers and business leaders to prepare for the resulting "skills shock" and to make informed decisions about investments and workforce development. The paper aims to address this by providing a more robust and predictive framework for understanding technology adoption.

Methods

The core of the paper's methodology is the Emerging Technology Adoption S-Curve (ETASC) model. This model is built upon a data-driven pipeline with four main stages:

  1. Data Curation: Gathering performance indicators from various sources like patent databases, scientific publications, and technological news for each country. This data is then cleaned and missing values are imputed.
  2. Feature Engineering: Extracting and weighing key factors that influence technology adoption. These include economic development, regulatory environment, cultural norms, and infrastructure. A novel method is used to assign weights based on relative importance.
  3. Clustering: A machine learning algorithm (specifically K-means) is used to group countries into clusters (fast, average, slow adopters) based on their weighted features. A scoring system is then applied to rank countries within these clusters.
  4. Adoption Modelling: This stage combines country rankings with industry adoption data to estimate the impact of emerging technologies. It utilizes a novel parametric mathematical adoption model, enriched with technology domain knowledge, to predict adoption curves. The mathematical model is based on a hyperbolic trigonometric function, a variation of the Bass model, which uses parameters (k1, k2, k3, k4) to describe the S-shaped adoption curve. These parameters are interpolated for each country-industry pair using the ranking scores.

Datasets

The paper utilizes data from several sources for country assessment and scoring:

  • World Bank: For indicators like mobile broadband subscriptions.
  • World Economic Forum (WEF): Annual reports and datasets, including over 100 indicators across 12 productivity drivers (pillars) for 130 countries.
  • Central Intelligence Agency (CIA) Factbook.
  • UNESCO.

The paper also mentions using data related to Robotic Process Automation (RPA) trends, including Google Trends, VC Funding, Start-ups, and Patents, as well as Deloitte's global RPA survey, to establish a base adoption curve.

Key Insights and Ideas

  • Global Indicators are Key: The model's strength lies in its ability to derive global indicators as features to predict technology adoption curves, offering a more nuanced view than previous models.
  • Country and Industry Specificity: Recognizing that adoption rates vary significantly by country and industry is crucial. The ETASC model explicitly accounts for this by analyzing country clusters and industry adoption rates.
  • Cognitive Factors Matter: The paper emphasizes the importance of cognitive processes (beliefs, attitudes, decision-making) in technology adoption, suggesting that understanding these can lead to better adoption strategies.
  • Data-Driven Prediction: The integration of a data-driven pipeline with machine learning and a parametric mathematical model offers a more precise and accurate prediction of technology adoption compared to purely theoretical models.
  • RPA as a Case Study: The use of RPA as an example demonstrates the model's capability to predict adoption timelines, identifying commercial availability, pilot project completion, and mass-scaled adoption phases.

Key Results

The paper successfully applies its machine learning approach to cluster countries into "fast," "average," and "slow" technology adoption groups. For instance, it identifies 28 countries as fast adopters, 57 as average, and 45 as slow. The model then uses these clusters and industry adoption rates to predict technology adoption curves. As a specific example, the paper predicts that for RPA, mass-scaled adoption will be reached in fast-adopting countries like Singapore or the United States by 2028. The model's parameters (k1-k4) are interpolated for specific country-industry pairs, allowing for tailored adoption curve predictions, as illustrated with examples for Singapore (Fast), Indonesia (Average), and Zimbabwe (Slow) in the health sector.

Limitations and Future Work

The authors acknowledge that the accuracy of their model's predictions depends on the accuracy and regular updates of country and industry data. They also note that more research is needed to provide further validation and accuracy for the basis adoption curves derived from external knowledge sources.

Future work directions include:

  • Integrating Generative AI and Industry 4.0: Exploring how these technologies can enhance the cognitive model, provide personalized recommendations, and automate tasks.
  • Investigating Impact on Bass Model Coefficients: Understanding how Generative AI and Industry 4.0 might influence the parameters of mathematical models like the Bass model.
  • Adapting the Cognitive Model: Extending the model to account for emerging technologies like Augmented Reality (AR) and Virtual Reality (VR).
  • Sensitivity Analysis and Simulation: Conducting further analysis to ensure the robustness and effectiveness of the proposed cognitive model.
  • Connecting to Recommender Systems: Establishing a link between the cognitive model and recommender systems to understand how individuals process recommendations and to identify potential biases.
  • Exploring Design Strategies: Investigating user-centered design, gamification, and persuasive design for increasing technology adoption.
  • Role of Trust: Examining how trust in technology, companies, and governments influences adoption decisions.
  • Applying the Model to Different Contexts: Testing the Cognitive Model for Technology Adoption in diverse fields like healthcare and education to assess its generalizability.

Summary updated

Abstract

The widespread adoption of advanced technologies, such as Artificial Intelligence (AI), Machine Learning, and Robotics, is rapidly increasing across the globe. This accelerated pace of change is drastically transforming various aspects of our lives and work, resulting in what is now known as Industry 4.0. As businesses integrate these technologies into their daily operations, it significantly impacts their work tasks and required skill sets. However, the approach to technological transformation varies depending on location, industry, and organization. However, there are no published methods that can adequately forecast the adoption of technology and its impact on society. It is essential to prepare for the future impact of Industry 4.0, and this requires policymakers and business leaders to be equipped with scientifically validated models and metrics. Data-driven scenario planning and decision-making can lead to better outcomes in every area of the business, from learning and development to technology investment. However, the current literature falls short in identifying effective and globally applicable strategies to predict the adoption rate of emerging technologies. Therefore, this paper proposes a novel parametric mathematical model for predicting the adoption rate of emerging technologies through a unique data-driven pipeline. This approach utilizes global indicators for countries to predict the technology adoption curves for each country and industry. The model is thoroughly validated, and the paper outlines highly promising evaluation results. The practical implications of this proposed approach are significant because it provides policymakers and business leaders with valuable insights for decision-making and scenario planning.
BibTeX
@article{sobhanmanesh2023cognitive,
  author = {Fariborz Sobhanmanesh and Amin Beheshti and Nicholas Nouri and Natalia Monje Chapparo and Sandya Raj and Richard A. George},
  title = {{A Cognitive Model for Technology Adoption}},
  journal = {Algorithms},
  year = {2023},
  doi = {10.3390/a16030155},
}
Figure