Data Driven Performance Metrics for
Smart Manufacturing
Funding Source: National Institute of Standards and Technology (NIST)
Objective
■ Engineers play a vital role in sustainability. Incorporating correct
performance metrics, big data, and smart manufacturing gives us the
ability to improve the
Manufacturing Energy Consumption
effectiveness and efficiency
25,000
of manufacturing processes.
trillion Btu
20,000
8,631
7,798
6,744
6,836
15,000
2,430
2,279
2,174
10,000
5,000
■ Conservation efforts have
been working, but
improvements are still
needed. Engineers have a
responsibility to design,
manufacture, and process
materials and products in
an ever more effective and
efficient manner in order
better use our limited
resources and reduce
negative impacts.
7,541
7,339
3,477
2,833
2,275
2,554
2,368
2,341
2,314
1,705
7,426
6,835
6,095
3,035
2,656
2,370
1,565
6,468
5,911
5,803
2,839
2,851
2,490
0
1991 (19,697 1994 (21,633 1998 (23,797 2002 (22,666 2006 (21,098 2010 (19,062
TBTU)
TBTU)
TBTU)
TBTU)
TBTU)
TBTU)
MECS Survey Year and Total Consumption for All Purposes
Other
Fuel Oil and LPG
Coal and Coal Coke
Natural Gas
Net Electricity
Source: U.S. Energy Information Administration.
Performance Metrics
■ Smart performance metrics
require continuous
improvement while
maintaining the integrity of the
process.
■ Reacting and adapting to
external and internal process
changes such as an increase
in production rate.
■ Analyze energy performance
metrics for a subtractive
manufacturing process.
Specific Energy
Process Time
■ Functional units are used
to frame the performance
metrics in relative and
comparable terms.
Methodology
Define the system for evaluation.
What quality and quantity of relevant variables are required.
Evaluate performance metrics.
Subject performance metrics to an internal or external change.
Can the performance metrics guide the system to a decision
which improves the performance metrics while maintaining the
integrity of the process.
Material
Removal Rate
■ Machine: Mori Seiki NVD 1500 DCG.
■ Tool: 5/16” solid carbide center cutting end mill.
■ Internal/External change: Increase in production from 10 to 12 and 17
parts per day.
■ Priority: maintain daily processing
schedule
■ Trial 3 is the business as usual case (BAU)
■ 12 parts/day: Trial 7 results in lower energy intensity while maintaining similar
daily processing time compared with BAU.
■ 17 parts/day: Trial 9 results in lower energy intensity while maintaining similar
daily processing time compared with BAU.
■ Tool damage occurs at Trial 9.
■ A limiting factor is required to balance the
continuous improvements that the smart
manufacturing system wants to make.
■ Machine load is chosen
Processing Time vs Energy
Intensity for 10, 12, and 17 Parts
with Changing Chip Load
18
Processing Time vs Energy Intensity for 10, 12, and
17 Parts with Changing in Incremental Depth of Cut
18
10 Parts/Day
12 Parts/Day
17 Parts/Day
14
12
10
Energy Intensity (Wh/part)
16
16
14
Trial 6 - 1mm
Increment
6
4
2
0
1500
2000
Total Processing Time (Minutes)
2500
Chip Load
Cutting
Feed Rate
Incremental Depth
(#)
(inch/tooth)
Speed (rpm)
(mm/min)
of Cut (mm)
1
0.0003
1750
53.34
1.5
2
0.0006
1750
106.68
1.5
3
0.0005
1750
88.9
1.5
4
0.0005
1500
76.2
1.5
5
0.0005
2000
101.6
1.5
6
0.0005
1750
88.9
1
7
0.0005
1750
88.9
2
8
0.0007
1750
124.46
1.5
9
0.001
1750
177.8
1.5
Aluminum
Chip Load
Cutting
Feed Rate
Incremental Depth
Trial (#)
(inch/tooth)
Speed (rpm)
(mm/min)
of Cut (mm)
1
0.0003
1750
53.34
1.5
2
0.0006
1750
106.68
1.5
3
0.0005
1750
88.9
1.5
4
0.0005
1500
76.2
1.5
5
0.0005
2000
101.6
1.5
■ The machine load, in conjunction with the specific energy and energy
intensity performance
Energy Intensity vs Load for Identical Processes Across Aluminum and Steel
20
metrics, gives the smart
18
manufacturing system the
16
- Trials 1
ability to improve the energy
14
- Trials 2
performance while
12
- Trials 3
maintaining the integrity of
10
- Trials 4
the process.
8
■ Energy intensity trends lower
with an increase in load.
12
10
■ Highly volatile which is where
smart manufacturing excels.
8
6
4
2
0
0
8
Steel Trial
Implementation and Conclusion
500
1000
1500
2000
2500
Total Processing Time (Minutes)
Trial 1 - Chip Load 0.0003 in/tooth
Trial 7 - 2mm
Increment
Trial 2 - Chip Load 0.0006 in/tooth
Trial 3 - 1.5mm
Increment
Trial 9 - Chip Load 0.001 in/tooth
Trial 3 - Chip Load 0.0005 in/tooth
Trial 8 - Chip Load 0.0007 in/tooth
10 Parts/Day
12 Parts/Day
17 Parts/Day
Load% of load felt versus maximum rated load for Mori Seiki
NVD 1500
Case Study Results
1000
Machine Load
■ Smart Manufacturing (SM) refers to intelligent, automated
manufacturing systems that leverage information technology to
enable cost-effective and efficient fabrication.
Chosen Energy Performance Metrics
500
1018
Steel
Case Study
■ How to evaluate the smartness of performance metrics.
0
Energy Intensity
■ Enabled by
improvements in,
and a reduction in
cost of, monitoring
equipment and
computing power.
■ Real-time data
capture to react
and adapt to
internal and
external changes
to the system.
■ According to prioritized
concerns, performance
metrics are chosen for
evaluation and
improvement.
■
■
■
■
■
■ Understand how effective current standard manufacturing
performance metrics are in analyzing smart manufacturing
processes.
Smart Manufacturing
■ All three sustainability pillars should be reviewed, and their
subsets, to understand what aspects need to be considered
during analysis.
Energy Intensity (Wh/part)
David A. Jagger
© 2014 LMAS
contact email: [email protected]
Introduction & Motivation
- Trials 5
6
4
Steel Trials 1-5
2
Aluminum Trials 1-5
0
8.00
9.00
10.00
11.00
12.00
13.00
14.00
Energy Intensity (Wh/part)
15.00
16.00
17.00
■ To evaluate and improve the energy performance of the process only a
small amount of very detailed data is required: Energy use, material
removed, process time, and the machine load.
■ A limiting factor is required for effective use of performance metrics in a
smart manufacturing system.
Berkeley
/
UNIVERSITY OF CALIFORNIA