-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
299 lines (253 loc) · 13.1 KB
/
index.html
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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>Autofunk, a fast and scalable framework for building formal models from production systems | William Durand</title>
<link rel="stylesheet" href="reveal.js/css/reveal.min.css">
<link rel="stylesheet" href="reveal.js/css/theme/simple.css" id="theme">
<link rel="stylesheet" href="reveal.js/lib/css/zenburn.css">
<link rel="stylesheet" href="font-awesome/css/font-awesome.min.css">
<link rel="stylesheet" href="css/fix-fontawesome-reveal-js.css">
<link rel="stylesheet" href="css/custom.css">
<script>
if (window.location.search.match(/print-pdf/gi)) {
document.write('<link rel="stylesheet" href="css/pdf.css" type="text/css">');
}
</script>
<!--[if lt IE 9]>
<script src="reveal.js/lib/js/html5shiv.js"></script>
<![endif]-->
</head>
<body>
<div class="reveal">
<div class="slides">
<section data-background="images/title-bg.jpg">
<h1 class="title">Autofunk, a fast and scalable framework for building formal models from [Michelin] production systems.</h1>
<em class="subtitle">Sébastien Salva, <u>William Durand</u> — July 1st, 2015 / DEBS'15</em>
</section>
<section data-background="images/chaine-des-puys.jpg"></section>
<section>
<section data-markdown>
<script type="text/template">
## Quick Tour @ Michelin
</script>
</section>
<section data-markdown>
<script type="text/template">
A **factory** is divided into several **workshops**,
<br>
one for each step of the manufacturing process.
</script>
</section>
<section data-markdown>
<script type="text/template">
A **production system** is composed of devices, production
machines, and one or more software to control them.
<br>
<i class="fa fa-info-circle"></i> In our case, we target a single workshop only.
</script>
</section>
<section data-markdown>
<script type="text/template">
Software exchange _information_ with points and machines
by sending and receiving **production events**.
<br>
```
17-Jun-2015 23:29:59.50|17011|MSG_IN [pid: 1] [nsec: 8] [point: 1] ...
17-Jun-2015 23:29:59.61|17021|MSG_OUT [pid: 1] [nsec: 8] [point: 3] ...
17-Jun-2015 23:29:59.70|17011|MSG_IN [pid: 2] [nsec: 8] [point: 2] ...
```
<small>A simple example of 3 events in a human readable format</small>
</script>
</section>
<section data-markdown>
<script type="text/template">
**Production events** are exchanged in a binary
format (custom protocols), through **centralized**
exchanging systems.
</script>
</section>
<section data-markdown>
<script type="text/template">
Each production event is tied to a **product** (e.g. tire),<br>identified
by a **product identifier** (`pid`).
<br>
Gathering all production events related to a product<br>allows
to retrieve **what happened** to it.
</script>
</section>
</section>
<section>
<section data-markdown>
<script type="text/template">
## Background
</script>
</section>
<section data-markdown>
<script type="text/template">
### Developement Teams POV
<br>
* 100+ applications running in **production**
* Not (fully) covered by tests
* Documentation most likely outdated
* MUST be maintained for ~20 years!
</script>
</section>
<section data-markdown>
<script type="text/template">
### Customers (Factories) POV
<br>
* Stability over anything else
* Maintenance periods are planned,<br>but rather long (> 1 week)
* 1h (unexpected) downtime = 50k $
</script>
</section>
<section data-markdown>
<script type="text/template">
**Testing** such production systems is **complex**,<br>
and takes a lot of time as it implies the physical<br>
devices, and there are **numerous behaviours**.
</script>
</section>
<section data-markdown>
<script type="text/template">
These behaviours could be formally described into a model.
<br>But writing such models is an **heavy task** and **error prone**.
<br>
<i class="fa fa-info-circle"></i> Not suitable for Michelin applications.
</script>
</section>
<section data-markdown>
<script type="text/template">
### Our Approach (1/2)
<br>
By leveraging the information carried by the
production events, we build **formal and exact
models** (STS) that describe functional behaviours
of a production system under analysis.
</script>
</section>
<section data-markdown>
<script type="text/template">
### Our Approach (2/2)
<br>
Michelin's exchanging systems **guarantee the order** in which the production events occured.
We now **capture the events directly into these systems** to avoid event loss, reordering, and/or
duplication of the production events.
</script>
</section>
</section>
<section>
<section data-markdown class="no-border">
<script type="text/template">
## The Big Picture
</script>
</section>
<section>
<h3>In Depth Autofunk</h3>
<p>
<img src="images/autofunk.png" class="no-border" />
</p>
</section>
<section data-markdown>
<script type="text/template">
### Autofunk
<br>
* Combines different fields: model inference,
<br>expert systems, and (now) machine learning
* Written in Java 8, reusing powerful
libraries<br>(e.g. [Spark](https://spark.apache.org), [Drools](http://www.drools.org/))
* More a **P**roof **o**f **C**oncept than a production-<br>ready tool
* To be open sourced (no ETA yet)
</script>
</section>
<section data-markdown class="no-border centered-table">
<script type="text/template">
### Experimentation
<br>
10 million production events (20 days)
<i class="fa fa-arrow-down"></i>
161,035 traces
<i class="fa fa-arrow-down"></i>
| `S` | `R(S)` |
|-----------------|----------------|
| 77,058 branches | 1,587 branches |
| 43,536 branches | 1,585 branches |
<br>
<i class="fa fa-info-circle"></i> 2 entry points here
It took 5 minutes to build the two models.
</script>
</section>
</section>
<section>
<section data-markdown>
<script type="text/template">
## Work In Progress
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Offline Passive Testing
<br>
* Inferred models are used as specifications
* Another set of traces is collected on a system<br>under test `SUT` (new or upgraded)
<br>
<i class="fa fa-arrow-down"></i>
<br>
<br>
**Does `SUT` conforms to the specifications?**
</script>
</section>
</section>
<section>
<section data-markdown>
<script type="text/template">
## Conclusion
<br>
* **Fast** and **efficient** technique to infer formal models
* The more production events, the better!
* But a few technical issues to tackle (memory<br>consumption for instance)
</script>
</section>
</section>
<section data-markdown>
<script type="text/template">
## Future Work
<br>
* Deploying Autofunk as a real solution (WIP)
* Offline passive testing (WIP)
* Online passive testing
</script>
</section>
<section data-markdown>
<script type="text/template">
## Thank You.
### Questions?
<p class="sponsors">
<img src="images/blaisepascal.png" class="no-border" />
<img src="images/limos.png" class="no-border" />
<img src="images/michelin.jpg" class="no-border" />
</p>
</script>
</section>
</div>
</div>
<script type="text/javascript" src="reveal.js/lib/js/head.min.js"></script>
<script type="text/javascript" src="reveal.js/js/reveal.min.js"></script>
<script type="text/javascript">
Reveal.initialize({
slideNumber: !window.location.search.match(/print-pdf/gi),
history: true,
theme: Reveal.getQueryHash().theme,
transition: Reveal.getQueryHash().transition || 'concave',
dependencies: [
{ src: 'reveal.js/lib/js/classList.js', condition: function() { return !document.body.classList; } },
{ src: 'reveal.js/plugin/markdown/marked.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'reveal.js/plugin/markdown/markdown.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'js/highlight.js', async: true, callback: function() { hljs.initHighlightingOnLoad(); } },
]
});
</script>
</body>
</html>