AI Adoption and Readiness in Manufacturing

HEIMBERGER, H., HORVAT, D., JÄGER, A. y SCHULTMANN, F., 2025. Exploring AI Adoption in Manufacturing: An Empirical Study on Effects of AI Readiness. International Journal of Production Economics, pp. 109733. ISSN 0925-5273.

• DOI 10.1016/j.ijpe.2025.109733.

• Quantitative Survey Study / Large-scale empirical analysis. Germany. Manufacturing (general) - applicable to Food Processing, Dairy, Meat. Full-scale (1,334 firms)

Technologies:

• AI applications (machine learning, pattern recognition).

• Self-learning algorithms with pattern recognition capabilities. Various levels from standalone to fully integrated production systems.

Industrial applications:

• Production process management

• Quality control

• Predictive maintenance

• Internal logistics management

Maturity and readiness

• Tech: "Majority of companies fall into medium (49%) or high (40%) technological readiness categories. Solid technological foundation established."

• Eco-fin: "Gap between AI readiness and actual adoption suggests apprehension towards implementation. 14% adoption rate despite higher readiness levels."

• Organisational: "High level organisational readiness less frequently achieved (28%). One in four manufacturers exhibit no (5%) or only low (20%) organisational readiness."

• Supply chain: "AI applications in internal logistics management (6% adoption). Supply chain integration limited but present."

• Legal framework: "Data security measures: 60% implement few measures, 33% various security measures. Regulatory compliance addressed through security protocols."

Results

• AI Adoption Rate

  • 14% of German manufacturing companies

  • Binary variable (0/1) across 4 application areas

  • Representative sample

• Technological Readiness

  • 89% medium-high level

  • Ordinal scale (0-4 levels)

  • Statistically significant

• Organisational

  • 75% medium-high level

  • Ordinal scale (0-4 levels)

  • p<0.05

• Combined Readiness Impact

  • 3x higher odds

  • Logistic regression OR=3.04

  • p<0.01

Implementation factors

Levers

• Data availability (82% basic-large), IT infrastructure (69% adequate), security measures (93% implemented)

• Clear ROI expectations for complex product manufacturers

• Training programmes (81% offered), innovation culture (73% present), R&D cooperation (57%)

• Industry sector influence (automotive 31% vs others ~14%)

Barriers

• Limited integration capabilities, legacy system constraints

• Unclear returns on investment, cost concerns for medium-sized firms

• Skills gaps (47% lack digital skills), resistance to change, cultural barriers

• Regulatory uncertainty, sector-specific constraints (chemicals/pharma 7% adoption)

KEY findings

• Finding 1: Despite relatively high AI readiness levels (89% technological, 75% organisational), actual adoption remains low (14%), indicating implementation hesitancy.

• Finding 2: Combined AI readiness significantly more predictive than individual dimensions - holistic approach essential for successful adoption.

• Finding 3: Structural characteristics crucial - product complexity provides 3x advantage, firm size shows U-shaped relationship.

• Finding 4: Organisational readiness emerges as more critical than technological readiness for actual implementation success.