科研 Skill
自驱动实验室闭环材料发现
high-throughput-autonomous-labs
Covers combinatorial synthesis, robotic experimentation. Use for autonomous or robotics-driven experimental campaigns, a less standardized domain.
SKILL.md
High-Throughput and Autonomous Materials Discovery
Overview
Autonomous/self-driving labs close the loop between synthesis, characterization, and decision-making — an algorithm (typically Bayesian optimization or a related active-learning method) selects the next experiment based on prior results, without waiting for a human to manually plan each iteration. This is a genuinely different workflow from every other skill in this repo, which assume a human plans each characterization/synthesis step; it's included because it's real and growing, but the tooling here is less standardized than characterization or DFT conventions.
When to use this skill
Activate when the request involves:
- High-throughput synthesis, combinatorial screening, autonomous lab, self-driving lab, active learning for materials discovery, Bayesian optimization (applied to synthesis/experiments)
- Terms: acquisition function, exploration-exploitation, closed-loop experimentation, design of experiments for screening
- "Plan a high-throughput screening campaign," "set up Bayesian optimization for synthesis parameters"
Core usage
Bayesian optimization for synthesis parameter selection
from skopt import gp_minimize
from skopt.space import Real
def objective(params):
"""Wraps the actual synthesis + characterization pipeline — in a real
autonomous lab, this triggers hardware and waits for a measured result;
here it's the interface point where this skill hands off to
synthesis-processing and the relevant characterization skill."""
temperature, precursor_ratio = params
measured_property = run_synthesis_and_characterize(temperature, precursor_ratio) # placeholder for the real pipeline
return -measured_property # gp_minimize minimizes; negate if maximizing the target property
search_space = [Real(400, 800, name="temperature_C"), Real(0.5, 2.0, name="precursor_ratio")]
result = gp_minimize(objective, search_space, n_calls=30, random_state=42)
best_params = result.x
Design-of-experiments screening (before optimization — establishing the response surface)
from scipy.stats import qmc
def latin_hypercube_screening_design(n_samples, parameter_bounds):
"""Space-filling design for initial screening — better coverage of the
parameter space than a naive grid for a given experiment budget, standard
practice before committing to a more targeted optimization campaign."""
sampler = qmc.LatinHypercube(d=len(parameter_bounds))
unit_samples = sampler.random(n=n_samples)
lower_bounds, upper_bounds = zip(*parameter_bounds)
return qmc.scale(unit_samples, lower_bounds, upper_bounds)
Validation & Pitfalls
Canonical references: Shahriari et al. (2016), "Taking the human out of the loop: A review of Bayesian optimization," Proceedings of the IEEE, for the general methodology; MacLeod et al. (2020), "Self-driving laboratory for accelerated discovery of thin-film materials," Science Advances, for a representative autonomous-lab implementation.
- Bayesian optimization's efficiency gains depend on the underlying response surface being reasonably smooth/well-behaved — a genuinely discontinuous or highly multimodal property landscape can mislead the surrogate model and the optimizer toward a local rather than global optimum. Don't assume convergence to a good result without checking whether the explored space and found optimum make physical sense.
- The measurement/characterization step feeding the optimization loop has its own noise and systematic error (see the relevant characterization skill's Validation & Pitfalls) — an optimization loop built on noisy or biased measurements can converge confidently to a wrong answer. The optimization algorithm doesn't know the measurement is unreliable; validate measurement quality independently, not just optimization convergence behavior.
- This field's tooling and conventions are less standardized than the rest of materials characterization — a workflow that works well for one lab's specific hardware/software stack often doesn't transfer directly to another's. Treat the code here as illustrating the underlying optimization concepts, not a drop-in pipeline; real autonomous-lab integration requires hardware/software-specific engineering this skill doesn't cover.
- A successful autonomous campaign optimizes the specific objective function defined — a poorly chosen objective (e.g. optimizing one property while ignoring a constraint like stability or cost) can produce a "successful" result that's practically useless. Define the objective function (including any constraints, as penalty terms or explicit constraint handling) carefully before running a campaign, not after seeing where it converges.
- Batch/parallel experimentation (common in robotic labs, running many experiments simultaneously rather than one at a time) requires optimization strategies designed for batch acquisition — a purely sequential Bayesian optimization strategy applied naively to batch selection can pick redundant (too-similar) experiments within a batch; use a batch-aware acquisition strategy when the lab setup actually runs experiments in parallel.
示例提示词
帮我设计一个贝叶斯优化引导的机器人合成闭环实验方案。
所属插件
安装该插件会一并获得同领域的 42 个 Skill / MCP。
同插件的其他科研 Skill
- 贝叶斯优化材料发现Covers Bayesian optimization for materials and chemistry design: surrogate models, acquisition functions, batch/constrained BO. Use when planning sample-efficient experiments or simulations.
- CALPHAD热力学计算工作流Covers CALPHAD computational thermodynamics: database selection, equilibrium/metastable calculations, binary/ternary phase diagrams. Use when predicting phase stability or solidification paths.
- 材料第一性原理与分子动力学模拟Routes first-principles and classical materials simulation (DFT, MD, ML potentials) to the right specialized workflow. Use as an entry point for any simulation-based materials task.
- 蠕变力学建模与验证Selects, calibrates, and validates creep deformation and rupture models under sustained thermomechanical load, covering rate prediction. Use when assessing long-term high-temperature components.
- 晶体塑性有限元建模Covers crystal plasticity and CPFE modeling of metals: slip systems for FCC/BCC/HCP, Schmid factors. Use when simulating texture- and orientation-dependent plastic deformation.
- 衍射结构表征Covers diffraction-based structure characterization (XRD, neutron, SAED): space group determination, Rietveld refinement, phase identification. Use for any diffraction-pattern analysis technique.
- EBSD织构分析Analyzes EBSD orientation maps for grain size, misorientation, phase, and texture, with pole figures and twin-boundary identification. Use for electron backscatter diffraction workflows.
- 电化学性能测试分析Covers electrochemical characterization: cyclic voltammetry, battery cycling analysis, EIS. Use for electrochemistry-specific data analysis distinct from general materials characterization.
- 金属疲劳寿命评估Builds, calibrates, and validates fatigue assessments for metals and components: stress-life, strain-life, multiaxial critical-plane. Use when predicting fatigue life or crack growth.
- 多物理场有限元建模Covers finite element modeling for coupled materials, mechanics, and transport problems: weak-form construction, boundary conditions, mesh choice. Use for multiphysics FEM simulations.
- 连续尺度有限元力学建模Covers continuum-scale mechanical, thermal, and multiphysics simulation with Abaqus/COMSOL conventions, mesh generation. Use for macroscale FEM consuming measured or computed material properties.
- 断裂力学评估与建模Selects, constructs, and validates fracture-mechanics assessments: LEFM checks, elastic-plastic fracture mechanics. Use when assessing cracked materials or components.
- 高斯过程材料回归建模Covers Gaussian process regression for materials and chemistry datasets: descriptor selection, kernel design. Use when a probabilistic surrogate model with uncertainty is needed.
- 材料数据库可复现访问Covers reproducible API access to materials databases (Materials Project, AFLOW, OQMD, NOMAD, COD, ICSD-style sources): authentication. Use when programmatically querying these databases.
- 材料数据库跨库查询Covers querying public materials databases beyond Materials Project: AFLOW, OQMD, NOMAD, JARVIS. Use when pulling or comparing data from these live, API-integrated resources.
- 材料数据标准与互操作Covers data interoperability standards for materials science: CIF crystal files, the OPTIMADE query API. Use when structuring or exchanging materials data between tools.
- 材料结构特征化Covers converting structures, compositions, and trajectories into ML feature vectors: Magpie-style elemental statistics. Use when building inputs for materials ML models.
- 材料科学图表绘制规范Covers publication-quality figure conventions for materials science plots: phase diagrams, pole figures, stress-strain curves. Use when the deliverable is a figure, not just an analysis result.
- 材料信息学性质预测Covers ML for materials property prediction and generative design: descriptor-based models, crystal GNNs (CGCNN/MEGNet-class). Use for materials informatics distinct from interatomic potentials.
- 材料生命周期评估Covers life-cycle assessment for materials and manufacturing routes: goal/scope definition, functional units, inventory construction. Use for environmental-impact studies of materials.
- 材料科学文献检索Covers formulating literature search queries for materials science questions and synthesizing/citing results responsibly. Requires a web search tool; it cannot substitute for one.
- 工程材料选型设计Selects engineering materials via material indices and Ashby charts, ranking candidates by performance per cost or mass. Use when picking a material class for a design constraint.
- 材料统计方法与实验设计Covers statistics shared across materials characterization and simulation: design of experiments. Use for cross-cutting statistical treatment of materials data.
- 材料结构与成分核心分析Covers core structure and composition analysis: reading/writing crystal structures, composition handling, symmetry analysis, phase diagrams. Use as a general-purpose structure-analysis toolkit.
常见问题
自驱动实验室闭环材料发现能做什么?
Covers combinatorial synthesis, robotic experimentation. Use for autonomous or robotics-driven experimental campaigns, a less standardized domain。
如何安装自驱动实验室闭环材料发现?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「自驱动实验室闭环材料发现」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
自驱动实验室闭环材料发现需要哪些权限?
该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
自驱动实验室闭环材料发现经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。