Category Theory for Semantic Data Interoperability Exchange and Consolidate Data for Agile PLM at Scale
Category Theory for Semantic Data InteroperabilityExchange and Consolidate Data for Agile PLM at Scale
San Francisco | New York City | Boston | Munich 2
© 2021 Conexus AI, Inc. – Confidential document.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 3
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Requirement
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 4
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 5
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
RequirementDesign
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 6
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 7
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Manufacturing Process Planning
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 8
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Control System Logic
Manufacturing Process Planning
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 9
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Control System Logic
Manufacturing Process Planning
Bill of Materials
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 10
Design
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 11
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 12
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 13
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 14
1
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 15
2
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 16
6
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 17
12
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 18
20
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 19
220
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 20
v2
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 21
v2
© 2021 Conexus AI, Inc.
~N2Data Integration Challenges
San Francisco | New York City | Boston | Munich
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 23Adapted from Breiner, S, Applied Category Theory workshop, National Institute Standards and Technology, 2017
Variety of Source Systems
Costs
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 25
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 26
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 27
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 28
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 29
??
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 30
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Control System Logic
Manufacturing Process Planning
Bill of Materials
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 32
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Control System Logic
Manufacturing Process Planning
Bill of Materials
Trying to standardize semantics globally sacrifice domain-specific semantic fidelity.
© 2021 Conexus AI, Inc.
Data Integration Challenges
San Francisco | New York City | Boston | Munich 33Adapted from Breiner, S, Applied Category Theory workshop, National Institute Standards and Technology, 2017
Sacrifice domain-specific semantic fidelity. VS
© 2021 Conexus AI, Inc.
Semantic Data InteroperabilityEnabled by Categorical Mathematics
Category Theory History
San Francisco | New York City | Boston | Munich 35
1890-1945 Foundational Ideas
Ideas leading to categories, functors,
and natural transformations
2011 OlogsDiscovery connecting categories to
database/ontology schemas
2019 Algebraic Property Graphs
APG published as an interoperable
enterprise knowledge graph meta-model
Uber u
ses A
CT to i
ntero
pera
te hu
ndre
ds
of tho
usan
ds of
data
sour
ces
2019
-Now
Ube
r Int
erop
erab
ility
Multipl
e pro
jects
launc
hing t
o
utiliz
e ACT t
o driv
e
engin
eerin
g sys
tem
inter
oper
abilit
y
Now
2018 Applied Category Theory
Applied Category Theory (ACT) expanded
more broadly to semantic IT interoperability
Applie
d Cate
gory
Theo
ry (A
CT) ut
ilized
commer
cially
acro
ss va
rious
indu
stries
to int
egra
te da
ta
2011
-2015
Vario
us A
pplic
ation
s
Catego
ry Th
eory
deve
loped
over
deca
des i
nto a
comple
te the
ory
1945
-1970
Ear
ly Dev
elopm
ent
1971 Categories for Working
Mathematician Published
Standard reference for Category
Theory
Dozen
s of c
ontrib
ution
s mad
e to f
urthe
r
deve
lop an
d exp
and C
atego
ry Th
eory
1971
-2011
Late
Develo
pmen
t
© 2021 Conexus AI, Inc.
Category Theory History
San Francisco | New York City | Boston | Munich 36
© 2021 Conexus AI, Inc.
David Spivak Ryan WisneskyMIT MIT
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 37
Variety of Source Systems
Costs
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 38
1
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 39
2
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 40
4
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 41
4 v 6
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 42
6
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 43
6 v 12
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 44
8
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 45
8 v 20
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 46
22
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 47
22 vs 220
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 48
v2
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 49
v2
© 2021 Conexus AI, Inc.
Foundationally Better Cost Scaling
San Francisco | New York City | Boston | Munich 50
v2
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 51
<<requirement>>Climb: One-engine-inoperative
FEDERAL AVIATION ADMINISTRATION (FAA) Title 14, Chapter 1, Subchapter C 25.121
…(a) Takeoff; landing gear extended… the steady gradient of column… not less than 0.5 percent… for four-engine airplanes
Perform simulation under conditions...
<<testcase>>Climb: One-engine-inoperativeTest Case
Requirement
Simulation
Design
Control System Logic
Manufacturing Process Planning
Bill of Materials
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 52
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 53
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 54
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 55
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 56
© 2021 Conexus AI, Inc.
New Data Integration Paradigm
San Francisco | New York City | Boston | Munich 57
© 2021 Conexus AI, Inc.
Required Inputs
● Data Schema● Data Records● Schema Links (as schema)● Record Links (as records)● Rules (declarative)
Automatically Compose Data Sources
San Francisco | New York City | Boston | Munich 58
© 2021 Conexus AI, Inc.
Automatically Compose Data Sources
San Francisco | New York City | Boston | Munich
Data, Concepts, Relationships & Rules AKA Records, Attributes, References & Constraints
© 2021 Conexus AI, Inc.
Exploiting Categorical Mathematics
● Automate composition of integrations● Merge data models, schemas and ontologies safely● Exchange and consolidate data● Enable forward and backward compatibility● Curate and add sources incrementally● Enforce global and domain-specific rules● Machine-verify integrity of integrations● Quantify implications of imperfect integrations● Interoperate between any kind of data structure
Conexus Enables New Paradigm
San Francisco | New York City | Boston | Munich 60
© 2021 Conexus AI, Inc.
Semantic Fabric
San Francisco | New York City | Boston | Munich 61
© 2021 Conexus AI, Inc. – Confidential document.
Semantic Fabric
San Francisco | New York City | Boston | Munich 62
© 2021 Conexus AI, Inc. – Confidential document.
Semantic Data InteroperabilityEnabled by Categorical Mathematics