LIGC Robotics / Design standard
Specimen · Phase 2 · For approval
Style specimen
The visual acceptance surface for the LIGC Robotics design standard. Every token, type role, rule, control and component the site will be built from, rendered once, at size, with its measured values beside it.
This is not a marketing page. Sample content illustrates layout and hierarchy; it asserts no customer, deployment, partnership or result. The page is not linked from the site and is excluded from search engines.
Design intent
An instrument for a capital decision, written by someone who has run a floor.
70%
Linear — structure
Near-black technical surfaces, hairline rules, controlled geometry, tight information hierarchy, interface-like data blocks, and severe restraint with colour.
30%
Anthropic — editorial
Warm off-white editorial bands, generous vertical air, long-form readability, a rationed serif for the argument, and calm human moments between technical sections.
Borrowed as discipline, never as template. No acid lime, no clay, no copied layout. If a decision makes the site feel more like generic AI SaaS, cyberpunk robotics, ecommerce or a startup template, it is rejected — however attractive it is.
Industrial intelligence · Executive credibility · Robotics precision
Near-black and warm white carry the site. One accent exists — LIGC Blue, sampled from the official emblem — and it is rationed: primary actions, the wordmark, section ticks, one editorial emphasis, and current state. Every ratio below is measured at build time from tokens.css against the surface the token is specified for.
The accent — LIGC Blue
Text
Light surfaces
Dark surfaces
Status — validation only
Accent budget — four uses per page
- 01 Primary buttons.
- 02 Section ticks — 2px × 40px on the section rule, major sections only.
- 03 One editorial emphasis: a single word, figure or process node.
- 04 Current state in navigation and the process flow.
If a fifth use appears, one of the first four was not important.
The one forbidden pairing
LIGC Blue is unusable as text or as a mark on a dark surface. The rule is demonstrated here rather than asserted: below is the measured ratio. On dark, the accent becomes --blue-on-dark — or stays a fill, which carries white type perfectly well.
Surface rhythm
A page alternates value deliberately. Two dark bands never sit adjacent. Carbon is the hero and the final call; between them at most two more dark bands. Warm marks the editorial and human moments.
Three families. Archivo speaks and acts; IBM Plex Mono names and measures; Source Serif 4 argues, and is rationed to two blocks per page. Weight is capped at 600 — hierarchy comes from size, space and rule, never from fat type. Sentence case throughout; uppercase is reserved for mono labels and button text.
Archivo — structure, voice and action
Before you buy a robot, know if it will actually work.
The Robotics Opportunity Assessment
What the assessment covers
Current-state process and labor baseline
Site readiness
We look at the actual work — the tasks, the floor, the volumes, the people — and tell you plainly whether a robot belongs in it.
Findings are written up and presented. The recommendation is one of four: proceed, pilot, investigate, or do not proceed. A recommendation not to automate is a valid outcome and is delivered as plainly as any other.
Sample content shown for layout only. It describes no customer and no completed engagement.
IBM Plex Mono — labels, indices and measured values
Recommendation / outcome
Ref 01 / A
Founding client · $1,500
El Paso, TX · 2026-09-06
Source Serif 4 — the argument, rationed
Two blocks per page, maximum. Never navigation, never UI, never labels, never buttons, never general body copy. A serif-heavy page is a failed page.
“Can a robot do it?” is not the question.
Type in context
Prose sets to a 68-character measure; the editorial serif to 60. Neither runs the full width of the grid, at any viewport.
A recommendation not to automate is a valid outcome
Most robotics conversations start with a machine. A vendor demonstrates a unit, the unit is impressive, and only afterwards does anyone examine the workflow, the facility, the labor model, the economics, the integration burden, the support path or the risk.
LIGC reverses that order. The operation is analysed first, and the equipment question is asked last — if it is asked at all. That sequence is the whole product.
A 1320px container with 40px desktop gutters, on a 12/8/4 column grid. Vertical air is the primary Anthropic borrow and is not compressed to fit more content: the large section rhythm is permitted at most twice per page.
Desktop ≥1024 · 12 columns · 40px gutter · 32px gap · 1320px max width
Tablet 768–1023 · 8 columns · 32px gutter · 24px gap · 768px max width
Mobile <768 · 4 columns · 20px gutter · 16px gap · 390px max width
Spacing scale
Section rhythm
Mobile · Tablet · Desktop
Every major section opens with a 1px rule and a 24px metadata band, then 40px of clear space, then content. Major sections carry a 2px × 40px accent tick flush left on that rule. Rules are always borders — never box-shadows, never pseudo-elements with a height. Radius is zero everywhere and the shadow budget is zero.
Section anatomy
- 01 1px top rule — the section boundary
- 02 2px × 40px accent tick — major sections only
- 03 24px metadata band — index and label left, reference right
- 04 40px clear space
- 05 Content on the grid
Rules & dividers
- Hairline — 1px --rule
- Major divider — 1px --rule-strong
- Section marker — 2px × 40px --blue, flush left on the top rule
- Row 01Row 02Row 03 Table row rule — 1px --rule between rows, none after the last
- Type on a carbon surface Hairline on dark — 1px --rule-dark
Radius & shadow
One radius token, value zero, applied to buttons, inputs, images and panels alike. This is a deliberate divergence from Linear’s subtle radii: it is what makes the geometry read as engineered rather than as an app. No shadow token exists — elevation is a value change and a hairline.
Primary is the blue fill with white type — 8.0:1, and identical on every surface, because the blue fill is how the eye finds the action. Secondary is a hairline outline that inverts. The text action carries the accent only on hover, in whichever variant the surface calls for. Hover always darkens. Radius 0, 48px tall, 44px minimum. Disabled uses surface and text tokens, never opacity.
On light
On dark
Every field has a visible, persistent mono label. Required is a mono REQUIRED label, not an asterisk. Controls are 48px, radius 0, 16px type so iOS does not zoom. Validation runs on blur and submit, never on keystroke, and errors are linked with aria-describedby. Choice controls are ruled rows, not chips or pills.
Enter the company name
One process is enough — the assessment covers a single primary workflow
A 3px left edge plus a mono label. Never a filled pill, never colour alone — the text always carries the meaning. Two status colours only: the site has no machine-status UI, and a third status hue would compete with the one accent.
The commercial workflow as a single ruled track: five nodes, mono indices, one line each. Not five icon cards, and no icons at all — the track carries the sequence. One node may take the accent as current state; that is the only accent permitted here. On mobile it becomes a vertical ruled spine, never a horizontal scroller.
On carbon — as it appears on the homepage
-
Discover
Qualify the operation and identify the workflow worth examining.
-
Assess
Walk the site, baseline the current process, labor and volumes.
-
Select
Research the relevant robot categories and compare viable candidates.
-
Coordinate
Requirements, quotes, demonstrations, site readiness and training.
-
Optimize
Track deployment issues and tune the process once it is running.
On paper — as it appears on an interior page
-
Discover
Qualify the operation and identify the workflow worth examining.
-
Assess
Walk the site, baseline the current process, labor and volumes.
-
Select
Research the relevant robot categories and compare viable candidates.
-
Coordinate
Requirements, quotes, demonstrations, site readiness and training.
-
Optimize
Track deployment issues and tune the process once it is running.
A real table with a pinned label column. Fit is an 8px square plus the literal words — never a green check and a red cross, never shape or colour alone. The indicators are monochrome: the accent is not spent on data. The table scrolls inside its own container; the page body never scrolls sideways.
Sample content — layout only
| Operating condition | Material transport | Food running | Floor cleaning |
|---|---|---|---|
| Repetitive, high-volume task | Good fit | Good fit | Good fit |
| Route changes several times a day | Poor fit | Good fit | Poor fit |
| Aisles narrower than 1.2 m | Poor fit | Poor fit | Good fit |
| Frequent human judgement required | Poor fit | Poor fit | Poor fit |
| Typical shift pattern | Two shifts | Peak hours | Overnight |
A 240px mono label column and a fluid value column, 48px rows, a hairline between rows and none after the last. An unknown value is a literal dash plus a note — never estimated, never quietly omitted. This is how the assessment scope is stated, and it is the reason nothing on the page can be vague.
Robotics Opportunity Assessment — scope. Founding client.
What the assessment includes
- Discovery
- Qualification call and workflow selection
- Site walkthrough
- On site, one primary workflow or application
- Baseline
- Current-state process, labor and volumes
- Feasibility
- Robotics feasibility analysis and relevant category research
- Comparison
- Candidate solution comparison where appropriate
- Economics
- Preliminary cost and payback scenarios
- Readiness
- Site-readiness considerations, risks and limitations
- Deliverable
- Written findings and an executive findings presentation
What it is not
- Engineering design
- Not included. LIGC does not engineer robot cells.
- Safety certification
- Not included. LIGC does not perform machine-safety certification.
- Guaranteed return
- Not offered. Scenarios are preliminary, not promises.
- Equipment sale
- Not a sales pitch. A recommendation not to proceed is a valid outcome.
Commercial
- Founding client price
- 1,500 USD
- Recommendation
- Proceed · Pilot · Investigate · Do not proceed
- Typical timeline
- — Stated at qualification once the workflow is known
The factors weighed beyond an hourly wage, as an itemised ruled schedule rather than a feature grid. Factors are NAMED, never QUANTIFIED: the component takes no value field, so there is nowhere for a fabricated number to go. No figures, no percentages, no payback months.
- 01 Overtime What the process actually costs at peak, not at plan.
- 02 Turnover The cost of recruiting and re-training for the role repeatedly.
- 03 Training Time to competence, and who is not doing their own job meanwhile.
- 04 Throughput Whether the constraint is the person or something upstream.
- 05 Injury and ergonomics Lifting, reach, repetition, and what that has already cost.
- 06 Scrap and rework Error rates in the current process and where they originate.
- 07 Downtime What stops when the process stops, and for how long.
- 08 Deployment cost Site readiness, power, connectivity, layout changes, integration.
- 09 Support cost Parts, service, spares and the response time actually available.
- 10 Total cost of ownership The whole picture over the life of the equipment.
The Anthropic 30%: a serif statement on the warm ground, offset from the grid, with air above and below. This is the site’s defining brand moment and the only place the operator’s voice is allowed through. One emphasis is permitted, set as a rule beneath the phrase — never a highlight or a fill.
The question
“Can a robot do it?” is not the question.
The better question is whether it should — and that is an operations and finance question long before it is a robotics one.
Technical possibility is cheap to establish. A vendor can almost always show you a machine that performs the motion. What that demonstration cannot tell you is whether the motion is the constraint, whether your floor supports the machine, what it costs to keep running, and what happens on the day it stops.
That is the gap LIGC works in. We analyse the operation first and treat the equipment question as the last one, not the first.
Maximum two serif blocks per page. This is one of them.
Header at rest — nav link: rest · hover (Safety) · current (Assessment)
Header after 24px of scroll — the rule darkens to --rule-strong. Nothing else changes.
Mobile header with the panel open — 390px frame
Wordmark
Set in the emblem’s own relationship — LIGC in LIGC Blue, ROBOTICS in ink — and it inverts on a dark surface without a second component. The official emblem is unchanged and remains the mark for the footer, favicon, social, print and documents; drop the file into public/brand/ and it replaces this lockup wherever the full mark belongs.
Header size — one line, Archivo 600, 15px, +0.04em
Stacked, 2×
Mobile is designed, not reflowed. The process flow becomes a vertical ruled spine. Tables scroll inside their own container with the label column pinned. A single 56px sticky action bar carries the primary action once the hero has scrolled away — the only fixed element besides the header. No horizontal overflow at 320px.
Hero at 390px — headline capped at four lines, price line visible
Robotics Opportunity Assessment
Before you buy a robot, know if it will actually work.
Founding client $1,500
Process flow at 390px — a vertical ruled spine, not five stacked cards
-
Discover
Qualify the operation and identify the workflow worth examining.
-
Assess
Walk the site, baseline the current process, labor and volumes.
-
Select
Research the relevant robot categories and compare viable candidates.
-
Coordinate
Requirements, quotes, demonstrations, site readiness and training.
-
Optimize
Track deployment issues and tune the process once it is running.
Sticky action bar — 56px, 1px top rule, one action. Shown here in flow.
Table at 390px — scrolls inside its own container, label column pinned
| Operating condition | Material transport | Food running | Floor cleaning |
|---|---|---|---|
| Repetitive, high-volume task | Good fit | Good fit | Good fit |
| Route changes several times a day | Poor fit | Good fit | Poor fit |
| Aisles narrower than 1.2 m | Poor fit | Poor fit | Good fit |
| Frequent human judgement required | Poor fit | Poor fit | Poor fit |
| Typical shift pattern | Two shifts | Peak hours | Overnight |
Four consecutive bands at full scale, so the system can be judged as a page rather than as parts: the carbon hero, the offer on paper, the warm editorial moment, and the final call. No two share a shape. The accent appears four times across the whole sequence.
Photography slot — reserved, empty by design until rights are cleared
Robotics Opportunity Assessment
Before you buy a robot, know if it will actually work.
LIGC Robotics evaluates your operation, identifies where robotics makes operational and financial sense, compares viable solutions, and helps you build a practical path to deployment.
How it runs
- 01 Discover
- 02 Assess
- 03 Select
- 04 Coordinate
- 05 Optimize
The offer
One workflow, examined properly, and a recommendation you can act on.
The assessment is a fixed-scope engagement on a single primary workflow. You are paying for a better decision, not for a robot.
What the assessment includes
- Discovery
- Qualification call and workflow selection
- Site walkthrough
- On site, one primary workflow or application
- Baseline
- Current-state process, labor and volumes
- Feasibility
- Robotics feasibility analysis and relevant category research
- Comparison
- Candidate solution comparison where appropriate
- Economics
- Preliminary cost and payback scenarios
- Readiness
- Site-readiness considerations, risks and limitations
- Deliverable
- Written findings and an executive findings presentation
What it is not
- Engineering design
- Not included. LIGC does not engineer robot cells.
- Safety certification
- Not included. LIGC does not perform machine-safety certification.
- Guaranteed return
- Not offered. Scenarios are preliminary, not promises.
- Equipment sale
- Not a sales pitch. A recommendation not to proceed is a valid outcome.
Commercial
- Founding client price
- 1,500 USD
- Recommendation
- Proceed · Pilot · Investigate · Do not proceed
- Typical timeline
- — Stated at qualification once the workflow is known
The recommendation is one of four
- Proceed
- Pilot
- Investigate
- Do not proceed
The question
“Can a robot do it?” is not the question.
The better question is whether it should — and that is an operations and finance question long before it is a robotics one.
Technical possibility is cheap to establish. A vendor can almost always show you a machine that performs the motion. What that demonstration cannot tell you is whether the motion is the constraint, whether your floor supports the machine, what it costs to keep running, and what happens on the day it stops.
That is the gap LIGC works in. We analyse the operation first and treat the equipment question as the last one, not the first.
Next step
Before you spend money on automation, make sure the operation supports it.
Mechanical, restrained, mostly absent. Nothing loops, nothing moves while the reader is reading, no transform exceeds 8px, one property at a time. prefers-reduced-motion: reduce removes every transition and animation on this page, and the process flow renders at its final state.
- --motion-fast 150 ms
- --motion-base 220 ms
- --ease-mech cubic-bezier(0.2, 0, 0, 1)
- Read the full scope Arrow translateX(3px) — Hover or focus the action
- Book a readiness call background-color · border-color · color — Hover the button — it darkens, never lightens
- 010203 First-paint reveal — opacity + 6px, max 3 items, 60ms stagger — Ran once on load. Reload to replay.
Binding. Listed as text, never demonstrated. Founder approval in writing is the only exception.
Structure
- endless equal-sized card grids
- the repeating headline / paragraph / three cards / CTA cadence
- two consecutive sections sharing a shape
- cards as default containment
- excessive rounded rectangles · badge pills · giant pill buttons
- centred H1 + centred paragraph + two pill buttons
- hero carousels · scroll-jacking · heavy parallax
Surface
- gradient blobs · mesh gradients · any decorative gradient
- glowing borders · neon · glassmorphism
- sparkles · ambient particles · floating UI cards
- heavy drop shadows — the shadow budget is zero
- giant gradient text
Imagery
- a humanoid robot face as the homepage gimmick
- sci-fi concept art · glowing robot brains · holograms
- circuit-board wallpaper · fake circuitry
- AI-generated machinery of any kind
- decorative robots unrelated to the service
- generic robot icon collections
- stock people pointing at screens
- any image without a recorded rights manifest
Content integrity
- fabricated clients, logos, testimonials or case studies
- “Trusted by” strips without real customers
- invented statistics, ROI results or deployment counts
- claimed certifications, authorisations or partnerships
- decorative dashboards depicting a product LIGC does not operate
- “unlock the future of automation” and every phrase like it
- “AI-powered” as generic marketing language
Technical
- component library aesthetic — no shadcn, DaisyUI, Flowbite, Preline
- CSS or utility-class framework
- motion library — CSS only
- third-party font requests at runtime
- client JS where CSS or static HTML suffices
- a literal hex, size or duration in a component stylesheet