Innovation Management GuideSAFE

🔬 A curated collection of 23,000+ agent skills for empirical research across 8 social science disciplines. | 精选 23,000+ AI Agent 技能库,覆盖8大社会科学学科的实证研究。CoPaper.AI 20分钟完成一篇可复现的规范实证论文,并支持用户上传 Skills。-- Maintained by CoPaper.AI from Stanford REAP.

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: innovation-management-guide
description: "Innovation metrics, R&D management research, and technology forecasting"
metadata:
  openclaw:
    emoji: "💡"
    category: "domains"
    subcategory: "business"
    keywords: ["innovation", "r-and-d", "technology-management", "bibliometrics", "diffusion", "strategy"]
    source: "wentor"
---

# Innovation Management Guide

A skill for conducting research on innovation management, technology strategy, and R&D performance. Covers innovation measurement, technology forecasting, diffusion modeling, patent-publication linkage, and bibliometric analysis of research portfolios.

## Innovation Measurement

### Key Innovation Metrics

| Metric | Definition | Data Source |
|--------|-----------|-------------|
| R&D intensity | R&D spending / Revenue | Annual reports, Compustat |
| Patent count | Granted patents per year | USPTO, EPO |
| Citation-weighted patents | Patents weighted by forward citations | PatentsView |
| New product revenue share | Revenue from products < 3 years old | Internal data |
| Time to market | Concept to commercial launch | Project records |
| Innovation efficiency | Revenue from new products / R&D spend | Combined internal data |

### Building an Innovation Scorecard

```python
import pandas as pd
import numpy as np

def compute_innovation_scorecard(firm_data: pd.DataFrame) -> pd.DataFrame:
    """
    Compute a multi-dimensional innovation scorecard for firms.
    firm_data columns: firm_id, rd_spend, revenue, patents_filed,
    patents_granted, citation_count, new_product_revenue, employees
    """
    scorecard = pd.DataFrame()
    scorecard["firm_id"] = firm_data["firm_id"]

    # Input metrics
    scorecard["rd_intensity"] = firm_data["rd_spend"] / firm_data["revenue"]
    scorecard["rd_per_employee"] = firm_data["rd_spend"] / firm_data["employees"]

    # Output metrics
    scorecard["patent_yield"] = (
        firm_data["patents_granted"] / (firm_data["rd_spend"] / 1e6)
    )
    scorecard["citation_impact"] = (
        firm_data["citation_count"] / firm_data["patents_granted"].clip(lower=1)
    )
    scorecard["new_product_share"] = (
        firm_data["new_product_revenue"] / firm_data["revenue"]
    )

    # Efficiency
    scorecard["innovation_efficiency"] = (
        firm_data["new_product_revenue"] / firm_data["rd_spend"]
    )

    # Normalize to percentile ranks within the sample
    for col in scorecard.columns[1:]:
        scorecard[f"{col}_rank"] = scorecard[col].rank(pct=True)

    # Composite score (equal weights)
    rank_cols = [c for c in scorecard.columns if c.endswith("_rank")]
    scorecard["composite_score"] = scorecard[rank_cols].mean(axis=1)

    return scorecard.sort_values("composite_score", ascending=False)
```

## Technology Diffusion Models

### Bass Diffusion Model

The Bass model is the foundational framework for forecasting technology adoption:

```python
from scipy.optimize import curve_fit

def bass_model(t: np.ndarray, p: float, q: float, m: float) -> np.ndarray:
    """
    Bass diffusion model for cumulative adoption.
    t: time periods (0, 1, 2, ...)
    p: coefficient of innovation (external influence)
    q: coefficient of imitation (internal influence)
    m: market potential (total eventual adopters)
    Returns cumulative adoption at each time period.
    """
    return m * (1 - np.exp(-(p + q) * t)) / (1 + (q / p) * np.exp(-(p + q) * t))

def bass_incremental(t: np.ndarray, p: float, q: float, m: float) -> np.ndarray:
    """Bass model incremental (new adopters per period)."""
    F = bass_model(t, p, q, m) / m
    f = (p + q * F) * (1 - F)
    return m * f

def fit_bass_model(adoption_data: np.ndarray) -> dict:
    """
    Fit Bass diffusion parameters to observed adoption data.
    adoption_data: cumulative adoption counts per period.
    """
    t = np.arange(len(adoption_data))
    try:
        popt, pcov = curve_fit(
            bass_model, t, adoption_data,
            p0=[0.01, 0.3, adoption_data[-1] * 2],
            bounds=([0, 0, adoption_data[-1]], [1, 2, adoption_data[-1] * 10]),
            maxfev=10000,
        )
        return {
            "p_innovation": round(popt[0], 6),
            "q_imitation": round(popt[1], 6),
            "m_potential": round(popt[2], 0),
            "peak_period": round(np.log(popt[1] / popt[0]) / (popt[0] + popt[1]), 1),
            "q_p_ratio": round(popt[1] / popt[0], 2),
        }
    except RuntimeError:
        return {"error": "convergence_failed"}
```

### Typical Bass Parameters by Technology Category

| Technology | p (innovation) | q (imitation) | q/p ratio |
|-----------|---------------|---------------|-----------|
| Consumer electronics | 0.01-0.03 | 0.3-0.5 | 10-50 |
| Enterprise software | 0.005-0.02 | 0.2-0.4 | 10-80 |
| Medical devices | 0.001-0.01 | 0.1-0.3 | 10-300 |
| Social media platforms | 0.03-0.10 | 0.5-0.8 | 5-25 |

## Bibliometric Analysis of R&D Portfolios

### Publication Portfolio Analysis

```python
def analyze_research_portfolio(publications: pd.DataFrame) -> dict:
    """
    Bibliometric analysis of an organization's research portfolio.
    publications columns: doi, title, year, journal, citations,
    fields (list), authors (list), affiliations (list)
    """
    # Publication trend
    annual_pubs = publications.groupby("year").size()

    # Citation impact
    citation_stats = {
        "total_citations": publications.citations.sum(),
        "mean_citations": publications.citations.mean(),
        "median_citations": publications.citations.median(),
        "h_index": compute_h_index(publications.citations.values),
    }

    # Research field distribution
    all_fields = []
    for fields in publications.fields:
        all_fields.extend(fields)
    field_dist = pd.Series(all_fields).value_counts().head(20)

    # Collaboration patterns
    collab_rate = publications.affiliations.apply(
        lambda x: len(set(x)) > 1
    ).mean()

    return {
        "total_publications": len(publicati

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (0)

No findings outside the package's declared scope.

Gates applied: no_behavioural_pass.