-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
757 lines (717 loc) · 46.4 KB
/
Copy pathindex.html
File metadata and controls
757 lines (717 loc) · 46.4 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width,initial-scale=1">
<meta name="description" content="Assembly2Intent — an interactive research prototype exploring functional engineering context from CAD assemblies.">
<meta name="theme-color" content="#07111f">
<title>Assembly2Intent — Functional Engineering Context from CAD Assemblies</title>
<style>
:root{
--bg:#07111f; --bg2:#0b1728; --panel:rgba(16,31,51,.78); --panel-solid:#102038;
--text:#eef6ff; --muted:#9db0c7; --line:rgba(181,207,235,.16);
--cyan:#56d7ff; --blue:#6f89ff; --violet:#a985ff; --green:#56e39f; --amber:#ffc86b;
--red:#ff7a8a; --shadow:0 24px 80px rgba(0,0,0,.28);
--radius:24px; --max:1240px;
}
body.light{
--bg:#f5f8fc; --bg2:#edf3fa; --panel:rgba(255,255,255,.86); --panel-solid:#fff;
--text:#0b1728; --muted:#526579; --line:rgba(29,60,91,.12);
--shadow:0 20px 60px rgba(24,47,73,.12);
}
*{box-sizing:border-box}
html{scroll-behavior:smooth}
body{
margin:0; color:var(--text); background:
radial-gradient(circle at 10% 10%,rgba(86,215,255,.10),transparent 30%),
radial-gradient(circle at 90% 0%,rgba(169,133,255,.12),transparent 32%),
linear-gradient(180deg,var(--bg),var(--bg2));
font-family:Inter,ui-sans-serif,system-ui,-apple-system,BlinkMacSystemFont,"Segoe UI",sans-serif;
line-height:1.55; min-height:100vh;
}
body::before{
content:""; position:fixed; inset:0; pointer-events:none; opacity:.16; z-index:-1;
background-image:linear-gradient(var(--line) 1px,transparent 1px),linear-gradient(90deg,var(--line) 1px,transparent 1px);
background-size:48px 48px; mask-image:linear-gradient(to bottom,black,transparent 82%);
}
a{color:inherit}
button,select{font:inherit}
.container{width:min(var(--max),calc(100% - 36px));margin:auto}
.topbar{
position:sticky;top:0;z-index:50;backdrop-filter:blur(20px);
background:color-mix(in srgb,var(--bg) 76%,transparent);border-bottom:1px solid var(--line)
}
.nav{height:70px;display:flex;align-items:center;justify-content:space-between;gap:24px}
.brand{display:flex;align-items:center;gap:12px;text-decoration:none;font-weight:800;letter-spacing:-.03em}
.brandmark{
width:34px;height:34px;border-radius:10px;background:linear-gradient(135deg,var(--cyan),var(--violet));
position:relative;box-shadow:0 10px 30px rgba(86,215,255,.22)
}
.brandmark:before,.brandmark:after{content:"";position:absolute;border:2px solid rgba(255,255,255,.85);border-radius:50%}
.brandmark:before{width:9px;height:9px;left:6px;top:7px}
.brandmark:after{width:9px;height:9px;right:6px;bottom:7px}
.navlinks{display:flex;gap:4px;align-items:center}
.navlinks a,.ghost-btn{
color:var(--muted);text-decoration:none;border:0;background:transparent;padding:9px 11px;border-radius:10px;cursor:pointer
}
.navlinks a:hover,.ghost-btn:hover{color:var(--text);background:var(--line)}
@media(max-width:850px){.navlinks a:nth-child(-n+4){display:none}}
.hero{padding:92px 0 54px;position:relative}
.hero-grid{display:grid;grid-template-columns:1.2fr .8fr;gap:42px;align-items:center}
@media(max-width:900px){.hero-grid{grid-template-columns:1fr}.hero{padding-top:62px}}
.eyebrow{
display:inline-flex;align-items:center;gap:9px;color:var(--cyan);font-size:.82rem;font-weight:750;
letter-spacing:.12em;text-transform:uppercase
}
.eyebrow:before{content:"";width:26px;height:1px;background:var(--cyan)}
h1{
font-size:clamp(3rem,7vw,6.7rem);line-height:.91;letter-spacing:-.07em;margin:18px 0 24px;
max-width:980px
}
.gradient-text{background:linear-gradient(95deg,var(--cyan),#d4e6ff 45%,var(--violet));-webkit-background-clip:text;color:transparent}
.lede{font-size:clamp(1.08rem,2vw,1.38rem);max-width:790px;color:var(--muted);margin:0 0 30px}
.actions{display:flex;flex-wrap:wrap;gap:12px}
.btn{
display:inline-flex;align-items:center;justify-content:center;gap:9px;min-height:46px;padding:0 17px;
border-radius:13px;border:1px solid var(--line);text-decoration:none;font-weight:700;cursor:pointer
}
.btn.primary{background:linear-gradient(135deg,var(--cyan),var(--blue));color:#05121e;border:0}
.btn.secondary{background:var(--panel);color:var(--text)}
.btn:hover{transform:translateY(-1px);box-shadow:0 12px 34px rgba(0,0,0,.12)}
.hero-panel,.card,.phase-card,.evidence-box,.future-card,.chart-card,.gallery-card{
background:var(--panel);border:1px solid var(--line);border-radius:var(--radius);box-shadow:var(--shadow);
backdrop-filter:blur(16px)
}
.hero-panel{padding:26px;position:relative;overflow:hidden}
.hero-panel:after{
content:"";position:absolute;width:250px;height:250px;border-radius:50%;right:-100px;top:-110px;
background:radial-gradient(circle,rgba(86,215,255,.20),transparent 65%)
}
.mini-title{font-size:.78rem;text-transform:uppercase;letter-spacing:.13em;color:var(--muted);font-weight:800}
.metric-grid{display:grid;grid-template-columns:repeat(2,1fr);gap:12px;margin-top:15px}
.metric{
padding:18px;border:1px solid var(--line);border-radius:17px;background:color-mix(in srgb,var(--panel-solid) 50%,transparent)
}
.metric b{display:block;font-size:1.78rem;letter-spacing:-.045em}
.metric span{color:var(--muted);font-size:.84rem}
.statusline{display:flex;gap:8px;align-items:center;margin-top:16px;color:var(--muted);font-size:.88rem}
.dot{width:9px;height:9px;border-radius:50%;background:var(--green);box-shadow:0 0 0 5px rgba(86,227,159,.10)}
.section{padding:78px 0}
.section-head{display:flex;justify-content:space-between;align-items:end;gap:26px;margin-bottom:30px}
.section-head h2{font-size:clamp(2.1rem,5vw,4rem);letter-spacing:-.055em;line-height:1;margin:0}
.section-head p{color:var(--muted);max-width:620px;margin:0}
@media(max-width:760px){.section-head{display:block}.section-head p{margin-top:16px}}
.story-switch{display:flex;gap:6px;padding:5px;border:1px solid var(--line);border-radius:13px;width:max-content;background:var(--panel)}
.story-switch button{border:0;background:transparent;color:var(--muted);padding:8px 12px;border-radius:9px;cursor:pointer;font-weight:700}
.story-switch button.active{background:var(--panel-solid);color:var(--text)}
.grid-3{display:grid;grid-template-columns:repeat(3,1fr);gap:16px}
.grid-2{display:grid;grid-template-columns:repeat(2,1fr);gap:18px}
@media(max-width:900px){.grid-3,.grid-2{grid-template-columns:1fr}}
.card{padding:24px}
.card h3,.phase-card h3,.future-card h3{margin:8px 0 8px;font-size:1.25rem;letter-spacing:-.025em}
.card p,.phase-card p,.future-card p{color:var(--muted)}
.pill{
display:inline-flex;align-items:center;gap:7px;border:1px solid var(--line);border-radius:999px;
padding:5px 9px;font-size:.75rem;font-weight:800;text-transform:uppercase;letter-spacing:.06em;color:var(--muted)
}
.pill.done{color:var(--green);border-color:rgba(86,227,159,.28);background:rgba(86,227,159,.06)}
.pill.planned{color:var(--amber);border-color:rgba(255,200,107,.28);background:rgba(255,200,107,.06)}
.pill.caution{color:var(--red);border-color:rgba(255,122,138,.28);background:rgba(255,122,138,.06)}
.architecture{
padding:28px;display:grid;grid-template-columns:repeat(5,1fr);gap:12px;align-items:center
}
.arch-node{
border:1px solid var(--line);background:var(--panel-solid);padding:18px;border-radius:17px;min-height:118px;
cursor:pointer;transition:.2s ease;position:relative
}
.arch-node:hover,.arch-node.active{border-color:var(--cyan);transform:translateY(-3px);box-shadow:0 18px 40px rgba(86,215,255,.10)}
.arch-node small{color:var(--cyan);font-weight:800;text-transform:uppercase;letter-spacing:.08em}
.arch-node strong{display:block;margin-top:8px}
.arch-arrow{text-align:center;color:var(--muted);font-size:1.5rem}
@media(max-width:950px){.architecture{grid-template-columns:1fr}.arch-arrow{transform:rotate(90deg)}}
.arch-detail{margin-top:14px;padding:16px 18px;border-left:3px solid var(--cyan);background:rgba(86,215,255,.05);border-radius:10px;color:var(--muted)}
.phase-tools{display:flex;gap:8px;flex-wrap:wrap;margin-bottom:20px}
.filter-btn{border:1px solid var(--line);background:var(--panel);color:var(--muted);padding:8px 12px;border-radius:10px;cursor:pointer}
.filter-btn.active{color:var(--text);border-color:var(--cyan);background:rgba(86,215,255,.07)}
.phase-list{display:grid;gap:16px}
.phase-card{padding:26px;display:grid;grid-template-columns:86px 1fr auto;gap:20px;align-items:start;transition:.2s}
.phase-card:hover{border-color:rgba(86,215,255,.35)}
.phase-card.hidden{display:none}
.phase-number{
width:70px;height:70px;border-radius:18px;display:grid;place-items:center;
background:linear-gradient(135deg,rgba(86,215,255,.12),rgba(169,133,255,.12));font-size:1.35rem;font-weight:900
}
.phase-card .phase-links{display:flex;gap:8px;flex-wrap:wrap}
.small-link{font-size:.83rem;color:var(--cyan);text-decoration:none}
.small-link:hover{text-decoration:underline}
@media(max-width:720px){.phase-card{grid-template-columns:60px 1fr}.phase-number{width:54px;height:54px}.phase-card>:last-child{grid-column:2}}
.exp-shell{overflow:hidden}
.exp-tabs{display:flex;gap:7px;flex-wrap:wrap;margin-bottom:18px}
.exp-tab{
border:1px solid var(--line);background:var(--panel);color:var(--muted);border-radius:12px;padding:10px 13px;cursor:pointer;font-weight:750
}
.exp-tab.active{background:var(--panel-solid);border-color:var(--cyan);color:var(--text)}
.exp-panel{display:none}
.exp-panel.active{display:block}
.result-grid{display:grid;grid-template-columns:.78fr 1.22fr;gap:18px}
@media(max-width:900px){.result-grid{grid-template-columns:1fr}}
.evidence-box{padding:24px}
.big-result{font-size:clamp(2.3rem,6vw,5rem);line-height:1;letter-spacing:-.06em;margin:9px 0}
.big-result small{font-size:.9rem;letter-spacing:0;color:var(--muted)}
.evidence-box ul{padding-left:20px;color:var(--muted)}
.chart-card{padding:22px}
.chart-head{display:flex;justify-content:space-between;gap:14px;align-items:center;margin-bottom:18px}
.metric-select{
border:1px solid var(--line);background:var(--panel-solid);color:var(--text);padding:8px 10px;border-radius:10px
}
.bar-chart{display:grid;gap:11px}
.bar-row{display:grid;grid-template-columns:minmax(105px,180px) 1fr 52px;gap:12px;align-items:center}
.bar-label{font-size:.84rem;color:var(--muted);overflow:hidden;text-overflow:ellipsis;white-space:nowrap}
.bar-track{height:13px;background:var(--line);border-radius:999px;overflow:hidden}
.bar-fill{height:100%;border-radius:999px;background:linear-gradient(90deg,var(--cyan),var(--violet));width:0;transition:width .65s cubic-bezier(.2,.7,.2,1)}
.bar-value{text-align:right;font-variant-numeric:tabular-nums;font-size:.84rem;font-weight:750}
.callout{
margin-top:18px;padding:16px 18px;border:1px solid rgba(255,200,107,.24);background:rgba(255,200,107,.06);
border-radius:14px;color:var(--muted)
}
.callout strong{color:var(--text)}
.dataset-grid{display:grid;grid-template-columns:repeat(4,1fr);gap:14px}
@media(max-width:1000px){.dataset-grid{grid-template-columns:repeat(2,1fr)}}
@media(max-width:600px){.dataset-grid{grid-template-columns:1fr}}
.gallery-card{overflow:hidden;cursor:pointer}
.gallery-card img{width:100%;height:180px;object-fit:cover;display:block;background:var(--panel-solid);transition:.25s}
.gallery-card:hover img{transform:scale(1.025)}
.gallery-card .gbody{padding:15px}
.gallery-card b{display:block;margin-bottom:4px}
.gallery-card span{color:var(--muted);font-size:.84rem}
.future-card{padding:22px;cursor:pointer;position:relative;overflow:hidden}
.future-card:before{
content:"";position:absolute;left:0;top:0;bottom:0;width:3px;background:linear-gradient(var(--cyan),var(--violet))
}
.future-detail{display:none;margin-top:15px;border-top:1px solid var(--line);padding-top:15px;color:var(--muted)}
.future-card.open .future-detail{display:block}
.criteria{display:grid;grid-template-columns:repeat(3,1fr);gap:10px;margin-top:12px}
.criteria div{border:1px solid var(--line);border-radius:12px;padding:11px;font-size:.82rem}
@media(max-width:700px){.criteria{grid-template-columns:1fr}}
.accordion{border-top:1px solid var(--line)}
.acc-item{border-bottom:1px solid var(--line)}
.acc-btn{
width:100%;text-align:left;border:0;background:transparent;color:var(--text);padding:18px 0;
display:flex;justify-content:space-between;gap:20px;cursor:pointer;font-weight:750
}
.acc-content{display:none;padding:0 0 18px;color:var(--muted);max-width:880px}
.acc-item.open .acc-content{display:block}
.evidence-table{width:100%;border-collapse:collapse;font-size:.9rem}
.evidence-table th,.evidence-table td{padding:12px 10px;border-bottom:1px solid var(--line);text-align:left;vertical-align:top}
.evidence-table th{color:var(--muted);font-size:.75rem;text-transform:uppercase;letter-spacing:.07em}
.evidence-table a{color:var(--cyan);text-decoration:none}
.repo-tree{
background:#030a12;color:#cce6ff;border:1px solid var(--line);border-radius:18px;padding:20px;overflow:auto;
font:13px/1.65 ui-monospace,SFMono-Regular,Menlo,Consolas,monospace
}
.discussion{
background:linear-gradient(135deg,rgba(86,215,255,.08),rgba(169,133,255,.08));
border:1px solid rgba(86,215,255,.25);border-radius:28px;padding:32px
}
.discussion h3{font-size:clamp(1.8rem,4vw,3rem);letter-spacing:-.04em;margin:0 0 14px}
.quote{
border-left:3px solid var(--cyan);padding-left:18px;color:var(--muted);font-size:1.05rem;margin:22px 0
}
.footer{padding:44px 0 64px;color:var(--muted);border-top:1px solid var(--line);margin-top:70px}
.footer-grid{display:flex;justify-content:space-between;gap:30px;flex-wrap:wrap}
.kbd{font-family:ui-monospace,monospace;border:1px solid var(--line);border-bottom-width:2px;border-radius:6px;padding:2px 6px;font-size:.78rem}
.modal{position:fixed;inset:0;z-index:100;display:none;place-items:center;background:rgba(0,0,0,.78);padding:24px}
.modal.open{display:grid}
.modal img{max-width:min(1100px,96vw);max-height:84vh;border-radius:18px;box-shadow:var(--shadow);background:white}
.modal-close{position:absolute;right:22px;top:18px;color:white;border:1px solid rgba(255,255,255,.25);background:rgba(0,0,0,.25);border-radius:50%;width:42px;height:42px;cursor:pointer}
.toast{
position:fixed;right:22px;bottom:22px;z-index:120;background:var(--panel-solid);border:1px solid var(--line);
border-radius:13px;padding:12px 15px;box-shadow:var(--shadow);opacity:0;transform:translateY(10px);pointer-events:none;transition:.25s
}
.toast.show{opacity:1;transform:translateY(0)}
.technical-only{display:none}
body.technical .technical-only{display:block}
body.technical .executive-only{display:none}
code{font-family:ui-monospace,SFMono-Regular,Menlo,monospace;color:var(--cyan)}
</style>
</head>
<body>
<header class="topbar">
<div class="container nav">
<a class="brand" href="#top"><span class="brandmark"></span><span>Assembly2Intent</span></a>
<nav class="navlinks" aria-label="Primary">
<a href="#research">Research</a>
<a href="#phases">Phases</a>
<a href="#experiments">Experiments</a>
<a href="#roadmap">Roadmap</a>
<a href="#evidence">Evidence</a>
<button class="ghost-btn" id="themeBtn" aria-label="Toggle theme">◐ Theme</button>
</nav>
</div>
</header>
<main id="top">
<section class="hero">
<div class="container hero-grid">
<div>
<div class="eyebrow">Research prototype · v0.1 outreach freeze</div>
<h1>From CAD geometry to <span class="gradient-text">functional engineering context.</span></h1>
<p class="lede">
Assembly2Intent explores a simple question with large implications:
can an AI system learn not only <em>what geometry exists</em>, but <em>why parts interact</em>—and eventually use that context to improve manufacturing drawings?
</p>
<div class="actions">
<a class="btn primary" href="#experiments">Explore the evidence →</a>
<a class="btn secondary" href="docs/EXPERIMENT_003_CROSS_ASSEMBLY_GENERALIZATION.md">Read final experiment</a>
<button class="btn secondary" id="copyCitation">Copy project citation</button>
</div>
</div>
<aside class="hero-panel">
<div class="mini-title">What has actually been measured</div>
<div class="metric-grid">
<div class="metric"><b>0.996</b><span>Phase 1 PR-AUC with assembly context</span></div>
<div class="metric"><b>0.960</b><span>Phase 1 F1, body-held-out</span></div>
<div class="metric"><b>0.723</b><span>Phase 2 macro-F1, rigid vs revolute</span></div>
<div class="metric"><b>0.619</b><span>Phase 3 best mean balanced accuracy</span></div>
</div>
<div class="statusline"><span class="dot"></span> Three experimental phases complete · 16-test suite with optional real-data integration tests</div>
</aside>
</div>
</section>
<section class="section" id="research">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Research question</div><h2>What does a mechanical interface <em>mean</em>?</h2></div>
<p>We deliberately separated physical contact, explicit constraints, and motion semantics instead of treating them as the same signal.</p>
</div>
<div class="story-switch" aria-label="Content depth">
<button class="active" data-story="executive">Executive view</button>
<button data-story="technical">Technical view</button>
</div>
<div class="grid-3" style="margin-top:22px">
<article class="card">
<span class="pill done">Evidence</span>
<h3>1 · Contact relation</h3>
<p>Which B-Rep faces or component pairs physically meet in an assembled design?</p>
<div class="technical-only"><code>CAD/B-Rep + occurrence transforms → interface features</code></div>
</article>
<article class="card">
<span class="pill done">Evidence</span>
<h3>2 · Constraint relation</h3>
<p>Which contacting parts are explicitly jointed rather than merely supporting, touching, or incidental?</p>
<div class="technical-only"><code>contact graph ≠ joint graph</code></div>
</article>
<article class="card">
<span class="pill done">Evidence</span>
<h3>3 · Motion semantics</h3>
<p>Can geometry hint whether a joint is rigid, revolute, or another mechanical relationship?</p>
<div class="technical-only"><code>interface composition + pose → joint semantics hypothesis</code></div>
</article>
</div>
<div class="architecture card" style="margin-top:18px" id="arch">
<div class="arch-node active" data-detail="Raw public CAD geometry: ShapeManager/ACIS text, STEP/OBJ bodies, B-Rep topology, occurrence transforms, and assembly metadata."><small>Input</small><strong>CAD assembly</strong></div>
<div class="arch-arrow">→</div>
<div class="arch-node" data-detail="Local topology and geometry are extracted without using the target labels: faces, edges, surface types, adjacency, mesh measurements, pose, and inter-body proximity."><small>Representation</small><strong>B-Rep + context</strong></div>
<div class="arch-arrow">→</div>
<div class="arch-node" data-detail="Completed experiments measure contact identification, explicit-joint inference, and rigid-vs-revolute semantics under progressively stricter validation."><small>Learning</small><strong>Interface semantics</strong></div>
</div>
<div class="arch-detail" id="archDetail">Raw public CAD geometry: ShapeManager/ACIS text, STEP/OBJ bodies, B-Rep topology, occurrence transforms, and assembly metadata.</div>
</div>
</section>
<section class="section" id="phases">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Program structure</div><h2>Three completed phases. Three planned.</h2></div>
<p>The first three phases are measured experiments. The next phases are explicit research hypotheses—not completed features.</p>
</div>
<div class="phase-tools">
<button class="filter-btn active" data-filter="all">All phases</button>
<button class="filter-btn" data-filter="done">Completed</button>
<button class="filter-btn" data-filter="planned">Planned</button>
</div>
<div class="phase-list">
<article class="phase-card" data-status="done">
<div class="phase-number">01</div>
<div>
<span class="pill done">Completed</span>
<h3>Physical interface identification</h3>
<p>Determine whether assembly context helps identify real contact faces compared with isolated geometry.</p>
<div class="phase-links"><a class="small-link" href="docs/EXPERIMENT_001_CONTACT_FACE_BASELINE.md">Experiment report ↗</a></div>
</div>
<div><b>PR-AUC 0.996</b><br><span style="color:var(--muted);font-size:.83rem">vs 0.781 geometry-only RF</span></div>
</article>
<article class="phase-card" data-status="done">
<div class="phase-number">02</div>
<div>
<span class="pill done">Completed</span>
<h3>Interface semantics within a design</h3>
<p>Separate physical contact from explicit joint intent, then test whether interface geometry carries rigid-vs-revolute information.</p>
<div class="phase-links"><a class="small-link" href="docs/EXPERIMENT_002_INTERFACE_SEMANTICS.md">Experiment report ↗</a></div>
</div>
<div><b>Macro-F1 0.723</b><br><span style="color:var(--muted);font-size:.83rem">3/4 revolute interfaces recovered</span></div>
</article>
<article class="phase-card" data-status="done">
<div class="phase-number">03</div>
<div>
<span class="pill done">Completed</span>
<h3>Cross-assembly generalization</h3>
<p>Train on two complete designs and test on a third unseen assembly. This intentionally exposed a strong domain-shift problem.</p>
<div class="phase-links"><a class="small-link" href="docs/EXPERIMENT_003_CROSS_ASSEMBLY_GENERALIZATION.md">Experiment report ↗</a></div>
</div>
<div><b>Balanced acc. 0.619</b><br><span style="color:var(--muted);font-size:.83rem">best mean rigid-vs-revolute transfer</span></div>
</article>
<article class="phase-card" data-status="planned">
<div class="phase-number">04</div>
<div>
<span class="pill planned">Planned hypothesis</span>
<h3>Functional-neighborhood representation</h3>
<p>Expand from the immediate contact patch to nearby B-Rep topology: bores, shafts, pins, holes, neighboring faces, repeated geometry, and the component assembly graph.</p>
</div>
<div><b>Goal</b><br><span style="color:var(--muted);font-size:.83rem">Improve unseen-design semantics</span></div>
</article>
<article class="phase-card" data-status="planned">
<div class="phase-number">05</div>
<div>
<span class="pill planned">Planned hypothesis</span>
<h3>Drawing-intelligence bridge</h3>
<p>Test whether learned functional context improves view selection, section/detail priorities, and which dimensions should be surfaced first.</p>
</div>
<div><b>Goal</b><br><span style="color:var(--muted);font-size:.83rem">Connect intent → drawing decisions</span></div>
</article>
<article class="phase-card" data-status="planned">
<div class="phase-number">06</div>
<div>
<span class="pill planned">Planned hypothesis</span>
<h3>Human-reviewed datum / GD&T and verification</h3>
<p>Use functional interfaces to propose—not autonomously finalize—datum candidates, critical relations, and a generate→verify→correct drawing loop.</p>
</div>
<div><b>Goal</b><br><span style="color:var(--muted);font-size:.83rem">Decision support, not blind automation</span></div>
</article>
</div>
</div>
</section>
<section class="section" id="experiments">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Interactive evidence</div><h2>Follow the result—and the failure.</h2></div>
<p>The story gets progressively harder: strong contact reconstruction, promising within-design semantics, then a strict transfer test that reveals what the representation is missing.</p>
</div>
<div class="exp-tabs">
<button class="exp-tab active" data-exp="exp1">Experiment 001</button>
<button class="exp-tab" data-exp="exp2">Experiment 002</button>
<button class="exp-tab" data-exp="exp3">Experiment 003</button>
</div>
<div class="exp-shell">
<section class="exp-panel active" id="exp1">
<div class="result-grid">
<article class="evidence-box">
<span class="pill done">Measured evidence</span>
<div class="mini-title" style="margin-top:16px">Contact-face identification</div>
<div class="big-result">0.996 <small>PR-AUC</small></div>
<p style="color:var(--muted)">Geometry + assembly context, Extra Trees, Leave-One-Body-Out on 111 visible B-Rep faces.</p>
<ul>
<li>50 true contact faces / 61 non-contact faces</li>
<li>48 TP, 59 TN, 2 FP, 2 FN at threshold 0.5</li>
<li>Geometry-only RF: PR-AUC 0.781</li>
<li>Hidden 10-face body detected and excluded before training</li>
</ul>
<a class="small-link" href="docs/EXPERIMENT_001_CONTACT_FACE_BASELINE.md">Open full methodology ↗</a>
</article>
<article class="chart-card">
<div class="chart-head"><div><b>Model comparison</b><div style="color:var(--muted);font-size:.82rem">Switch metrics</div></div>
<select class="metric-select" data-chart="chart1">
<option value="pr">PR-AUC</option><option value="f1">F1</option><option value="bal">Balanced accuracy</option>
</select>
</div>
<div class="bar-chart" id="chart1"></div>
<div class="callout"><strong>Interpret carefully.</strong> Autodesk contact labels are proximity-based. This validates assembly-interface reconstruction, not general engineering intent.</div>
</article>
</div>
</section>
<section class="exp-panel" id="exp2">
<div class="result-grid">
<article class="evidence-box">
<span class="pill done">Measured evidence</span>
<div class="mini-title" style="margin-top:16px">Contact ≠ joint ≠ motion</div>
<div class="big-result">0.723 <small>macro-F1</small></div>
<p style="color:var(--muted)">Best grouped rigid-vs-revolute result within one real assembly.</p>
<ul>
<li>34 contacting component-pair interfaces</li>
<li>18 also explicitly jointed; 16 had no joint</li>
<li>14 rigid / 4 revolute</li>
<li>Best model recovered 3 of 4 revolute interfaces</li>
</ul>
<a class="small-link" href="docs/EXPERIMENT_002_INTERFACE_SEMANTICS.md">Open full methodology ↗</a>
</article>
<article class="chart-card">
<div class="chart-head"><div><b>Motion semantics</b><div style="color:var(--muted);font-size:.82rem">Grouped within-design evaluation</div></div>
<select class="metric-select" data-chart="chart2">
<option value="macro">Macro-F1</option><option value="bal">Balanced accuracy</option><option value="revRecall">Revolute recall</option><option value="revPR">Revolute PR-AUC</option>
</select>
</div>
<div class="bar-chart" id="chart2"></div>
<div class="callout"><strong>Key distinction:</strong> physical contact does not automatically encode engineering intent. Joint construction geometry can also differ from actual contact geometry.</div>
</article>
</div>
</section>
<section class="exp-panel" id="exp3">
<div class="result-grid">
<article class="evidence-box">
<span class="pill caution">Strict transfer test</span>
<div class="mini-title" style="margin-top:16px">Leave-One-Assembly-Out</div>
<div class="big-result">0.619 <small>balanced accuracy</small></div>
<p style="color:var(--muted)">Best mean rigid-vs-revolute cross-assembly result. The performance drop is the most useful finding of Phase 3.</p>
<ul>
<li>3 independent real Autodesk assemblies</li>
<li>122 physically contacting occurrence-pair samples</li>
<li>32 rigid + 16 revolute jointed contacts</li>
<li>Same semantic label appears with different local geometry across designs</li>
</ul>
<a class="small-link" href="docs/EXPERIMENT_003_CROSS_ASSEMBLY_GENERALIZATION.md">Open full methodology ↗</a>
</article>
<article class="chart-card">
<div class="chart-head"><div><b>Cross-assembly motion transfer</b><div style="color:var(--muted);font-size:.82rem">Mean over unseen-design folds</div></div>
<select class="metric-select" data-chart="chart3">
<option value="bal">Balanced accuracy</option><option value="f1">F1</option><option value="pr">PR-AUC</option>
</select>
</div>
<div class="bar-chart" id="chart3"></div>
<div class="callout"><strong>Research implication:</strong> the model needs a wider functional neighborhood. A revolute mechanism may be encoded by a nearby bore/shaft/pin even when the immediate contact patch is planar.</div>
</article>
</div>
</section>
</div>
</div>
</section>
<section class="section" id="datasets">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Real public CAD</div><h2>Evaluated on independent Autodesk assemblies.</h2></div>
<p>The public repository does not redistribute Autodesk CAD files. Each assembly can be obtained independently from the official Fusion 360 Gallery Assembly Dataset for reproduction under Autodesk's Dataset License.</p>
</div>
<div class="dataset-grid">
<article class="card"><span class="pill done">Phase 01</span><h3>53060_3cef188a</h3><p>111 visible B-Rep faces · 25 contact pairs. Used for the contact-interface benchmark.</p></article>
<article class="card"><span class="pill done">Phase 02/03</span><h3>54181_f02c9b19</h3><p>34 contacting occurrence-pair samples · 14 rigid · 4 revolute. Used for within-design semantics and cross-design transfer.</p></article>
<article class="card"><span class="pill done">Phase 03</span><h3>55931_84bfa135</h3><p>35 contacting occurrence-pair samples · 10 rigid · 3 revolute · additional joint types.</p></article>
<article class="card"><span class="pill done">Phase 03</span><h3>53232_f8760054</h3><p>53 contacting occurrence-pair samples · 8 rigid · 9 revolute. One of the strongest domain-shift checks.</p></article>
</div>
<div class="callout"><strong>Reproduce locally:</strong> follow <a class="small-link" href="docs/REPRODUCIBILITY.md">the reproducibility guide</a> to obtain the official data directly from Autodesk and rerun the experiments.</div>
</div>
</section>
<section class="section" id="roadmap">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Future research</div><h2>What Phase 3 tells us to build next.</h2></div>
<p>These are proposals for discussion—not completed claims. Click each card to inspect the hypothesis, implementation path, and success criterion.</p>
</div>
<div class="grid-3">
<article class="future-card">
<span class="pill planned">Phase 04</span><h3>Functional neighborhood graph</h3>
<p>Represent more than the contact patch.</p>
<div class="future-detail">
<b>Hypothesis</b><p>Motion intent becomes more transferable when the model sees nearby bores, pins, shafts, holes, adjacent faces, and component-graph context.</p>
<div class="criteria"><div><b>Build</b><br>B-Rep graph + assembly graph</div><div><b>Compare</b><br>contact-only vs neighborhood</div><div><b>Success</b><br>clear Phase 3 transfer gain</div></div>
</div>
</article>
<article class="future-card">
<span class="pill planned">Phase 05</span><h3>Drawing decision bridge</h3>
<p>Connect learned function to drafting priorities.</p>
<div class="future-detail">
<b>Hypothesis</b><p>Functional-interface embeddings can improve which views, section cuts, and dimensions deserve priority over purely geometric heuristics.</p>
<div class="criteria"><div><b>Build</b><br>ranking model</div><div><b>Evaluate</b><br>historical drawings</div><div><b>Success</b><br>better top-K drafting decisions</div></div>
</div>
</article>
<article class="future-card">
<span class="pill planned">Phase 06</span><h3>Human-in-the-loop verification</h3>
<p>Suggest critical relations, then verify.</p>
<div class="future-detail">
<b>Hypothesis</b><p>Engineering context can support datum/GD&T candidates and a generate→verify→correct workflow without delegating final engineering responsibility to a black box.</p>
<div class="criteria"><div><b>Build</b><br>confidence + rationale</div><div><b>Review</b><br>engineer approval</div><div><b>Success</b><br>fewer missed critical relations</div></div>
</div>
</article>
</div>
</div>
</section>
<section class="section" id="limitations">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Research hygiene</div><h2>What this prototype does <em>not</em> prove.</h2></div>
<p>These limitations are part of the result. They are also why the future phases are technically interesting.</p>
</div>
<div class="accordion">
<div class="acc-item open">
<button class="acc-btn">Contact detection is not engineering intent <span>−</span></button>
<div class="acc-content">Autodesk contact labels are derived from assembled geometric proximity. The near-perfect Experiment 001 result validates interface reconstruction, not a general semantic understanding of design intent.</div>
</div>
<div class="acc-item">
<button class="acc-btn">Within-design semantics did not fully transfer <span>+</span></button>
<div class="acc-content">Experiment 002 was promising, but Experiment 003 showed that the same joint type can be expressed with different immediate contact geometry across designs. That is a domain-shift problem, not a reason to hide the result.</div>
</div>
<div class="acc-item">
<button class="acc-btn">The dataset is still small for semantic claims <span>+</span></button>
<div class="acc-content">Phase 3 contains 122 contacting pair samples across three assemblies. It is enough to expose failure modes, but not enough to claim a broadly general mechanical-intent model.</div>
</div>
<div class="acc-item">
<button class="acc-btn">No autonomous GD&T or tolerance decisions are claimed <span>+</span></button>
<div class="acc-content">Datum/GD&T and tolerance intelligence remain future human-reviewed research directions. The current prototype does not make validated manufacturing-tolerance decisions.</div>
</div>
</div>
</div>
</section>
<section class="section" id="evidence">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">Traceability</div><h2>Every major claim links back to evidence.</h2></div>
<p>This makes the page suitable for technical review: results, predictions, code, and methodology can all be inspected directly in the repository.</p>
</div>
<div class="card" style="overflow:auto">
<table class="evidence-table">
<thead><tr><th>Claim / artifact</th><th>Status</th><th>Evidence</th></tr></thead>
<tbody>
<tr><td>Phase 1 contact-face benchmark</td><td><span class="pill done">Measured</span></td><td><a href="docs/EXPERIMENT_001_CONTACT_FACE_BASELINE.md">Report</a> · <a href="results/real/contact_53060_3cef188a/ml/model_comparison.csv">Model comparison CSV</a></td></tr>
<tr><td>Phase 2 interface semantics</td><td><span class="pill done">Measured</span></td><td><a href="docs/EXPERIMENT_002_INTERFACE_SEMANTICS.md">Report</a> · <a href="results/real/semantic_54181_f02c9b19/motion_type_model_comparison.csv">Motion results</a></td></tr>
<tr><td>Phase 3 unseen-design transfer</td><td><span class="pill done">Measured</span></td><td><a href="docs/EXPERIMENT_003_CROSS_ASSEMBLY_GENERALIZATION.md">Report</a> · <a href="results/real/cross_assembly_phase3/motion_cross_assembly.csv">Cross-assembly results</a></td></tr>
<tr><td>ShapeManager/B-Rep parser</td><td><span class="pill done">Implemented</span></td><td><a href="src/assembly2intent/smt_brep.py">Parser source</a> · <a href="tests/test_smt_brep.py">Tests</a></td></tr>
<tr><td>Functional-neighborhood GNN / graph model</td><td><span class="pill planned">Planned</span></td><td>Phase 4 hypothesis; no performance claim yet.</td></tr>
<tr><td>Drawing / GD&T intelligence</td><td><span class="pill planned">Planned</span></td><td>Phase 5–6 research direction; no production claim yet.</td></tr>
</tbody>
</table>
</div>
</div>
</section>
<section class="section" id="github">
<div class="container">
<div class="section-head">
<div><div class="eyebrow">GitHub-ready</div><h2>Built to be inspected, rerun, and extended.</h2></div>
<p>Place this <code>index.html</code> at the repository root and enable GitHub Pages from the main branch. All evidence links are relative, so the site travels with the repository.</p>
</div>
<div class="grid-2">
<div class="repo-tree">Assembly2Intent/
├── index.html ← this interactive site
├── README.md
├── PROJECT_PLAN.md
├── src/assembly2intent/
│ ├── smt_brep.py
│ ├── baseline.py
│ └── metrics.py
├── scripts/
│ ├── train_interface_semantics.py
│ └── train_cross_assembly_generalization.py
├── data/real/
├── results/real/
├── docs/
│ ├── EXPERIMENT_001_...
│ ├── EXPERIMENT_002_...
│ └── EXPERIMENT_003_...
└── tests/</div>
<div class="card">
<span class="pill done">Reproducible</span>
<h3>Suggested GitHub presentation</h3>
<p>Keep the repository public-facing description conservative: <strong>“Research prototype exploring functional engineering context from public CAD assemblies.”</strong></p>
<p>Enable GitHub Pages: <span class="kbd">Settings</span> → <span class="kbd">Pages</span> → Deploy from branch → <span class="kbd">main / root</span>.</p>
<p>Autodesk CAD assets are intentionally excluded from this public package. Reviewers can obtain the official Dataset directly from Autodesk and reproduce the experiments locally using the included guide.</p>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container discussion">
<div class="eyebrow">DraftAid discussion</div>
<h3>The question is no longer “can AI see the CAD?”</h3>
<p class="lede" style="margin-bottom:10px">The more interesting question is whether an engineering representation can learn which geometric relationships matter—and use that context to make better downstream drawing decisions.</p>
<div class="quote">“The prototype suggests that assembly context is powerful for reconstructing interfaces, while cross-design semantic transfer requires a richer functional neighborhood than the immediate contact patch.”</div>
<div class="actions">
<a class="btn primary" href="docs/OUTREACH_FREEZE_v0.1.md">Read outreach freeze →</a>
<a class="btn secondary" href="PROJECT_PLAN.md">Open research roadmap</a>
</div>
</div>
</section>
</main>
<footer class="footer">
<div class="container footer-grid">
<div><b style="color:var(--text)">Assembly2Intent v0.1</b><br>Early research prototype · Prepared by Mohammed Ali</div>
<div>Evaluated using the Autodesk Fusion 360 Gallery Dataset; raw Autodesk CAD is not redistributed here.<br>Not affiliated with or endorsed by Autodesk or DraftAid.</div>
</div>
</footer>
<div class="modal" id="imgModal" aria-hidden="true"><button class="modal-close" aria-label="Close image">×</button><img alt="Expanded assembly"></div>
<div class="toast" id="toast">Copied.</div>
<script>
const DATA = {"exp1":[{"name":"Geometry + Assembly \u00b7 Extra Trees","short":"Assembly + ET","f1":0.96,"pr":0.996,"bal":0.964,"kind":"assembly"},{"name":"Geometry + Assembly \u00b7 Random Forest","short":"Assembly + RF","f1":0.96,"pr":0.994,"bal":0.964,"kind":"assembly"},{"name":"Geometry + Assembly \u00b7 Logistic","short":"Assembly + LR","f1":0.899,"pr":0.84,"bal":0.908,"kind":"assembly"},{"name":"Proximity Rule","short":"Proximity","f1":0.889,"pr":0.826,"bal":0.898,"kind":"rule"},{"name":"Geometry Only \u00b7 Random Forest","short":"Geometry + RF","f1":0.552,"pr":0.781,"bal":0.633,"kind":"geometry"},{"name":"Geometry Only \u00b7 Extra Trees","short":"Geometry + ET","f1":0.633,"pr":0.695,"bal":0.671,"kind":"geometry"},{"name":"Geometry Only \u00b7 Logistic","short":"Geometry + LR","f1":0.54,"pr":0.635,"bal":0.581,"kind":"geometry"}],"exp2joint":[{"name":"Contact-count \u2265 2 rule","short":"Simple rule","f1":0.762,"pr":0.651,"bal":0.694},{"name":"Random Forest","short":"RF","f1":0.579,"pr":0.684,"bal":0.524},{"name":"Extra Trees","short":"ET","f1":0.562,"pr":0.68,"bal":0.594},{"name":"Logistic Regression","short":"LR","f1":0.5,"pr":0.592,"bal":0.535}],"exp2motion":[{"name":"Extra Trees","short":"ET","macro":0.723,"bal":0.768,"revRecall":0.75,"revPR":0.619},{"name":"Logistic Regression","short":"LR","macro":0.679,"bal":0.679,"revRecall":0.5,"revPR":0.427},{"name":"Random Forest","short":"RF","macro":0.379,"bal":0.393,"revRecall":0.0,"revPR":0.359}],"exp3joint":[{"name":"Contact-count \u2265 2 rule","short":"Simple rule","f1":0.568,"pr":0.51,"bal":0.614},{"name":"Logistic Regression","short":"LR","f1":0.471,"pr":0.52,"bal":0.555},{"name":"Extra Trees","short":"ET","f1":0.449,"pr":0.529,"bal":0.558},{"name":"Random Forest","short":"RF","f1":0.433,"pr":0.551,"bal":0.535}],"exp3motion":[{"name":"Invariant \u00b7 Logistic","short":"Invariant LR","f1":0.509,"pr":0.436,"bal":0.619},{"name":"Invariant \u00b7 Extra Trees","short":"Invariant ET","f1":0.317,"pr":0.531,"bal":0.458},{"name":"Invariant \u00b7 Random Forest","short":"Invariant RF","f1":0.167,"pr":0.484,"bal":0.522},{"name":"Full \u00b7 Extra Trees","short":"Full ET","f1":0.167,"pr":0.463,"bal":0.549},{"name":"Full \u00b7 Logistic","short":"Full LR","f1":0.121,"pr":0.398,"bal":0.404},{"name":"Full \u00b7 Random Forest","short":"Full RF","f1":0.111,"pr":0.414,"bal":0.51}],"assemblies":[{"id":"54181_f02c9b19","contacts":34,"jointed":18,"rigid":14,"revolute":4,"other":0},{"id":"55931_84bfa135","contacts":35,"jointed":16,"rigid":10,"revolute":3,"other":3},{"id":"53232_f8760054","contacts":53,"jointed":17,"rigid":8,"revolute":9,"other":0}]};
const $ = (q,root=document)=>root.querySelector(q);
const $$ = (q,root=document)=>[...root.querySelectorAll(q)];
function toast(msg){
const t=$("#toast"); t.textContent=msg; t.classList.add("show");
clearTimeout(window.__toast); window.__toast=setTimeout(()=>t.classList.remove("show"),1800);
}
// Theme
const savedTheme=localStorage.getItem("a2i-theme");
if(savedTheme==="light") document.body.classList.add("light");
$("#themeBtn").addEventListener("click",()=>{
document.body.classList.toggle("light");
localStorage.setItem("a2i-theme",document.body.classList.contains("light")?"light":"dark");
});
// Executive / technical
$$("[data-story]").forEach(btn=>btn.addEventListener("click",()=>{
$$("[data-story]").forEach(b=>b.classList.remove("active")); btn.classList.add("active");
document.body.classList.toggle("technical",btn.dataset.story==="technical");
}));
// Architecture explorer
$$(".arch-node").forEach(node=>node.addEventListener("click",()=>{
$$(".arch-node").forEach(n=>n.classList.remove("active")); node.classList.add("active");
$("#archDetail").textContent=node.dataset.detail;
}));
// Phase filter
$$(".filter-btn").forEach(btn=>btn.addEventListener("click",()=>{
$$(".filter-btn").forEach(b=>b.classList.remove("active")); btn.classList.add("active");
const f=btn.dataset.filter;
$$(".phase-card").forEach(card=>card.classList.toggle("hidden",f!=="all" && card.dataset.status!==f));
}));
// Experiment tabs
$$(".exp-tab").forEach(btn=>btn.addEventListener("click",()=>{
$$(".exp-tab").forEach(b=>b.classList.remove("active")); btn.classList.add("active");
$$(".exp-panel").forEach(p=>p.classList.remove("active"));
$("#"+btn.dataset.exp).classList.add("active");
}));
function renderBars(el, rows, metric){
el.innerHTML="";
const max=Math.max(...rows.map(r=>r[metric]||0),1);
rows.forEach((r,i)=>{
const row=document.createElement("div"); row.className="bar-row";
row.innerHTML=`<div class="bar-label" title="${r.name}">${r.short||r.name}</div>
<div class="bar-track"><div class="bar-fill"></div></div>
<div class="bar-value">${(r[metric]??0).toFixed(3)}</div>`;
el.appendChild(row);
requestAnimationFrame(()=>setTimeout(()=>$(".bar-fill",row).style.width=((r[metric]||0)/max*100)+"%",i*45));
});
}
renderBars($("#chart1"),DATA.exp1,"pr");
renderBars($("#chart2"),DATA.exp2motion,"macro");
renderBars($("#chart3"),DATA.exp3motion,"bal");
$$(".metric-select").forEach(sel=>sel.addEventListener("change",()=>{
const id=sel.dataset.chart;
const source=id==="chart1"?DATA.exp1:id==="chart2"?DATA.exp2motion:DATA.exp3motion;
renderBars($("#"+id),source,sel.value);
}));
// Future cards
$$(".future-card").forEach(card=>card.addEventListener("click",()=>card.classList.toggle("open")));
// Accordions
$$(".acc-btn").forEach(btn=>btn.addEventListener("click",()=>{
const item=btn.closest(".acc-item"); item.classList.toggle("open");
btn.querySelector("span").textContent=item.classList.contains("open")?"−":"+";
}));
// Assembly gallery modal
const modal=$("#imgModal");
$$(".gallery-card").forEach(card=>card.addEventListener("click",()=>{
$("img",modal).src=card.dataset.img; modal.classList.add("open"); modal.setAttribute("aria-hidden","false");
}));
$(".modal-close").addEventListener("click",()=>{modal.classList.remove("open");modal.setAttribute("aria-hidden","true")});
modal.addEventListener("click",e=>{if(e.target===modal){modal.classList.remove("open");modal.setAttribute("aria-hidden","true")}});
document.addEventListener("keydown",e=>{if(e.key==="Escape")modal.classList.remove("open")});
// Copy citation
$("#copyCitation").addEventListener("click",async()=>{
const text="Mohammed Ali. Assembly2Intent v0.1: Functional Engineering Context from CAD Assemblies. Research prototype using public Autodesk Fusion 360 Gallery assembly data, 2026.";
try{await navigator.clipboard.writeText(text);toast("Project citation copied");}
catch{toast("Copy unavailable — select citation from page source");}
});
</script>
</body>
</html>