Li Group RIGOR Self-Driving Discovery Figure

Status: Visual concept only — not product requirements, implementation status, physical qualification, or evidence that every pictured instrument is already integrated.

This brief maintains the broad RIGOR target vision as a science-first story: Li Group's existing data, theory, AI, and experimental capabilities become a governed discovery engine shared by digital catalysis, batteries, and hydrogen storage. The image labels current platform foundations and future physical qualification separately. Software module names remain visible but subordinate to the scientific actions and outcomes.

Assets

• Style reference: rigor-agent-first-narrative-v2.png
• Original: li-group-rigor-self-driving-discovery.png
• Previous result: li-group-rigor-self-driving-discovery-v2.png
• Current result: li-group-rigor-self-driving-discovery-v3.png
• Format: landscape PNG, 16:9, 1672 × 941 px

The v3 result uses v2 as an edit target. It retains the discovery thesis while separating the final Operation gate, current software scope, target outcomes, and future physical qualification.

Production Prompt

Use case: scientific-educational
Asset type: flagship 16:9 graphical abstract for Li Group presentations,
proposals, and RIGOR documentation

Input image role: edit target. Preserve the layout, all scientific capital
cards, domain lanes, central action loop, module labels, laboratory illustration,
outcome staircase, connectors, and premium Nature/Cell editorial character:
ivory-white background, deep navy typography, restrained isometric laboratory,
teal autonomous loop, coral human-attention cue, solid warm-gold evidence flow,
generous whitespace, crisp thin connectors, and professional scientific tone.

PRIMARY REQUEST

Maintain the target vision of Li Group's digital-materials capabilities growing
into a self-driving discovery ecosystem through RIGOR. Make the current
software/simulated foundation and future qualified physical loop explicit. The
scientific transformation remains the protagonist.

Add one compact outlined badge above the title, rendered exactly once:
“TARGET VISION”

TITLE — render exactly once:
"From Digital Materials to a Self-Driving Discovery Ecosystem"

SUBTITLE — replace the existing subtitle and render exactly once:
"Scientific capital, governed experiments, and qualified physical systems in one learning loop."

COMPOSITION

Tell one clear left-to-right story in three zones.

LEFT ZONE

Title exactly "Scientific Capital".

Show four compact stacked cards:

"DATA"
"DigCat · DigBat · DigHyd"

"THEORY"
"Descriptors · Microkinetics"

"AI"
"Agents · ML Potentials"

"EXPERIMENT"
"Synthesis · Characterization"

Connect all four cards to one slim neutral navy collector rail. From that
shared rail, branch into three parallel thin domain lanes that enter the RIGOR
Discovery Engine together:

"DigCat" — warm coral-orange
"DigBat" — cobalt blue
"DigHyd" — emerald green

The construction must make it unmistakable that every scientific domain uses
all four layers of scientific capital. Do not map one domain to only one card.

CENTER ZONE

Create a visually dominant circular scientific engine titled exactly:
"RIGOR Discovery Engine"

Arrange seven large scientific actions clockwise around one refined isometric
hybrid laboratory:

"Mission"
"Strategy"
"Design"
"Rehearsal"
"Execute"
"Interpret"
"Learn"

Add only these smaller secondary module labels beneath their scientific action:

under Strategy: "ASCEND"
under Design: "Composer"
under Execute: "PACE · LabFlow · LabBridge"
under Interpret: "PRISM"

The isometric laboratory should include recognizable, uncluttered cues for
automated synthesis, liquid handling, electrochemical testing, spectroscopy or
microscopy, battery cycling, and hydrogen-sorption testing. Include one
respectful human technician as a planned laboratory resource.

Add one small dashed coral exception branch with exactly:
"Needs you"
"Attention · Human Task"

It is not part of the normal route.

Show Rehearsal as a translucent teal digital-twin laboratory with a checkmark
before live execution. Predicted or simulated states are translucent teal;
committed measurements and evidence are solid warm gold.

Add a compact warm-orange checkpoint between Rehearsal and Execute:
"Operation Gate"
"exact input · current catalog"

This gate represents the final Operation-creation boundary and remains visually
distinct from the digital twin.

Enclose the central physical laboratory in a dashed warm-gold boundary labelled
"TARGET PHYSICAL LAB". It remains a future qualification target.

Add a subtle warm-gold feedback fan from Learn to the shared Scientific Capital
collector, showing that committed evidence updates data, theory, AI, and
experimental practice.

Around the central loop add one slim rail with this exact text:
"Policy · Calibration · Uncertainty · Provenance · Audit · Safe-stop"

RIGHT ZONE

Title exactly "Research at a Higher Level".

Add a small label above the staircase:
"TARGET OUTCOMES"

Create an ascending four-step staircase, receiving the three domain lanes:

"Better Candidates"
"Robust Materials"
"Design Principles"
"Self-Improving Ecosystem"

Use concise scientific icon cues only: activity, selectivity, and capacity for
Better Candidates; stability, manufacturability, and scale for Robust
Materials; causal mechanisms and transferable descriptors for Design
Principles; reusable data, models, agents, and autonomous campaigns for the
Self-Improving Ecosystem.

From the highest step, draw one strong, elegant warm-gold arrow back to
Scientific Capital. Make the meaning unmistakable: each campaign creates both
a material result and a reusable knowledge asset.

SCOPE RAIL

Add a slim two-part rail along the lower edge without covering the existing
policy rail. Render these exact labels:

"CURRENT SOFTWARE SCOPE  agent-first software · simulation · algorithmic analysis · governed human work"
"FUTURE QUALIFICATION  physical instruments · physical effects"

Use a solid teal sample for CURRENT SOFTWARE SCOPE and a dashed warm-gold sample
for FUTURE QUALIFICATION. The central physical laboratory remains an intended
endpoint, not an implementation or qualification claim.

VISUAL HIERARCHY

- The center loop occupies about 55 percent of the canvas.
- Scientific action labels and DigCat, DigBat, and DigHyd are primary.
- RIGOR module names are readable but clearly secondary.
- All text is horizontal, large, crisp, correctly spelled, and contained.
- Use flat vector-infographic rendering plus the refined isometric laboratory.
- Keep the ivory background and abundant negative space.

TEXT CONSTRAINTS

Use only the exact title, subtitle, TARGET VISION badge, zone titles, card text,
lane labels, seven action labels, four module labels, Needs you,
TARGET OUTCOMES, the safety/provenance rail, the two-part scope rail, and the
four outcome labels specified above. English only. Do not invent extra text,
acronyms, numbers, citations, logos, or captions.

AVOID

No generic AI brain as the main symbol, cloud database, dashboard UI, dark
background, neon cyberpunk, cartoon robots, paper thumbnails, raw code, GPIO,
MQTT, device SDK details, crossing arrows, decorative molecules, fake
citations, watermark, tiny illegible text, duplicated labels, or one-to-one
mapping between a domain lane and a scientific-capital card.

Review Checklist

• DigCat, DigBat, and DigHyd all receive data, theory, AI, and experiment inputs.
• The main loop reads clockwise as Mission, Strategy, Design, Rehearsal, Execute, Interpret, Learn.
• Rehearsal precedes live execution and remains visually distinct from evidence.
• The Operation Gate follows Rehearsal and precedes Execute.
• Planned human work is visible, while Needs you remains exceptional.
• Committed evidence updates all four scientific-capital layers.
• The outcome ladder progresses from candidates to a self-improving ecosystem.
• Software names support the science narrative instead of dominating it.
• TARGET VISION and TARGET OUTCOMES remain legible.
• The scope rail separates current platform foundations from future physical qualification.

Scientific Mapping Sources

• Li Group research interests
• Li Group code and databases
• Li Group publication list

For normative RIGOR behavior, consult the Agent-First System Design PRD.