Intelligent machine tools
Methods and systems for generating sensor data, wherein the sensor data includes measured transit time information for items flowing through a work process, accessing a set of control data for one or more machine tool workstations and accounting data for the work process, calculating a standard deviation of the measured transit time information, calculating an achievable minimum WIP for at least one of the workstations using current values of workstation performance parameters, receiving input including: information identifying process improvement projects and corresponding predictive performance parameter values, and information identifying resources available for process improvement, for the at least one of the workstations: determining an achievable minimum WIP using the corresponding predictive performance parameter values, determining a reduction in minimum WIP based on the difference between the achievable minimum WIP for the current performance parameter values and the predictive performance parameter values, and determining one or more process efficiency improvements.
1. A system comprising:
a collection of machine tool workstations used in a work process, wherein each workstation is associated with a set of performance parameters and one or more of the machine tool workstations comprise (i) a cutting tool, (ii) a drilling machine, (iii) a lathe, (iv) a boring machine or (v) a grinding machine;
one or more data sensors, the one or more data sensors:
in data communication with each other, and
performing operations comprising:
generating a set of sensor data, wherein the sensor data includes measured transit time information for items flowing through the work process,
accessing a set of production control data for each workstation and a set of accounting data for the work process,
calculating a standard deviation of the measured transit time information, and
calculating an achievable minimum work in progress (WIP) for each workstation using current values of workstation performance parameters, comprising calculating an achievable minimum WIP for each workstation using
W
MIN
≅
SQm
2
(
1
-
X
-
m
(
P
~
/
(
1
-
M
)
)
)
+
Q
wherein S represents the workstation setup time, Q represents the workstation number of different part numbers, m represents the units produced by the workstation per unit time, X represents the workstation scrap percentage, {tilde over (P)} represents the workstation machining time per unit, and M represents the machine downtime;
receiving input including (i) information identifying process improvement projects and corresponding predictive performance parameter values, and (ii) information identifying resources available for process improvement;
for each workstation (i) determining an achievable minimum WIP using the corresponding predictive performance parameter values, and (ii) determining a reduction in minimum WIP based on the difference between the achievable minimum WIP for the current values of performance parameters and the predictive performance parameter values; and
determining one or more process efficiency improvements associated with the determined reductions in minimum WIP, the accounting data and the calculated standard deviation.
2. The system of claim 1 , wherein the one or more sensors in data communication with each other perform operations further comprising implementing one or more of the determined process efficiency improvements.
3. The system of claim 2 , wherein:
one or more of the determined process efficiency improvements comprises increasing a number of units produced per unit time by a cutting tool in the collection of machine tool workstations used in the work process; and
implementing one or more of the determined process efficiency improvements comprises adjusting the feed rate of the cutting tool.
4. The system of claim 3 , wherein adjusting the feed rate of the cutting tool comprises adjusting one or more of (i) a desired surface finish of an item processed by the cutting tool, (ii) power available to the cutting tool, (iii) threads per inch (TPI) for threading tools included in the cutting tool, or (iv) rigidity of the cutting tool.
5. The system of claim 1 , wherein the performance parameters include one or more of setup time, number of different part numbers, units produced per unit time, scrap percentage, and machining time per unit, and machine downtime.
6. The system of claim 1 , wherein the production control data is data relating to one or more of the performance parameters.
7. The system of claim 1 , wherein the accounting data for the work process is data relating to overhead costs and labor costs for the work process.
8. The system of claim 1 , wherein determining one or more process efficiency improvements associated with the determined reductions in minimum WIP, the accounting data and the calculated standard deviation comprises:
determining a total WIP prior to a process efficiency improvement based on the set of sensor data;
defining one or more histograms of K bins that represent, for one or more respective randomly selected items of WIP flowing through the work process, probability distributions that assign a probability to the velocities of the items;
for each combination of determined reductions in minimum WIP for one or more corresponding workstations:
subtracting the combined one or more determined reductions in minimum WIP for one or more corresponding workstations from the determined total WIP prior to process efficiency improvement;
determining a process efficiency improvement using:
process efficiency improvement=$μ (L+O)B σ τB √{square root over (2)} D ( {circumflex over (p)} WA |{circumflex over (p)} WB )
wherein $μ (L+O)B represents labor and overhead expended per month prior to process efficiency improvement; σ τB represents the calculated standard deviation; D represents a relative entropy
D
(
p
^
WA
|
p
^
WB
)
=
∑
j
=
1
K
p
^
WAj
log
p
^
WAj
p
^
WBj
,
and
p
^
WAj
=
1
/
2
(
n
Bj
b
_
j
/
W
B
)
,
p
^
WBj
=
1
/
2
(
n
Aj
b
_
j
/
W
A
)
,
wherein W B represents the total WIP prior to process efficiency improvement, W A represents the total WIP prior to process efficiency improvement minus the combined one or more determined reductions in minimum WIP for one or more corresponding workstations, K is a number of histogram bins, n Bj b j represents an amount of WIP prior to process efficiency improvement in the j-th bin, and n Aj b j represents an amount of WIP prior to process efficiency improvement minus the combined one or more determined reductions in minimum WIP for one or more corresponding workstations in the j-th bin.
9. The system of claim 8 , wherein {circumflex over (p)} WA i and {circumflex over (p)} WB i are distributed according to the Maxwell-Boltzmann distribution.
10. The system of claim 8 , further comprising a thermodynamic model configured to determine the one or more process efficiency improvements for the process, the process efficiency improvements being derived from thermodynamic principles.
11. The system of claim 10 , wherein the thermodynamic model includes a conversion factory β that is given by
(
1
β
)
=
$μ
(
L
+
O
)
B
2
σ
τ
2
wherein $μ (L+O)B represents labor and overhead expended per month prior to process efficiency improvement, and σ τB represents the standard deviation of the measured transit time information prior to process improvement.
12. The system of claim 10 , wherein the process efficiency improvement is derived from thermodynamic free energy.
13. The system of claim 10 , wherein the thermodynamic model is derived from manufacturing entropy.
14. The system of claim 13 , wherein the manufacturing entropy is derived from Little's Law and comprises a Shannon entropy and an entropy of scale.
15. The system of claim 14 , wherein the thermodynamic model is derived from maximizing the Shannon entropy.
16. The system of claim 1 , wherein calculating an achievable minimum work in progress (WIP) for each workstation comprises calculating an achievable minimum WIP for each workstation using
W
MIN
≅
SQm
2
(
1
-
X
-
m
(
P
~
/
(
1
-
M
)
)
)
+
Q
+
m
τ
t
wherein S represents the workstation setup time, Q represents the workstation number of different part numbers, m represents the units produced by the workstation per unit time, X represents the workstation scrap percentage, {tilde over (P)} represents the workstation machining time per unit, M represents the machine downtime, and τ t represents the average time WIP is in transit between workstations.
17. A computer implemented method, comprising:
generating a set of sensor data, wherein the sensor data includes measured transit time information for items flowing through a work process;
accessing a set of production control data for one or more machine tool workstations and a set of accounting data for the work process, wherein one or more of the machine tool workstations comprise (i) a cutting tool, (ii) a drilling machine, (iii) a lathe, (iv) a boring machine or (v) a grinding machine;
calculating a standard deviation of the measured transit time information;
calculating an achievable minimum work in progress (WIP) for at least one of the one or more workstations using current values of workstation performance parameters, comprising calculating an achievable minimum WIP for each workstation using
W
MIN
≅
SQm
2
(
1
-
X
-
m
(
P
~
/
(
1
-
M
)
)
)
+
Q
wherein S represents the workstation setup time, Q represents the workstation number of different part numbers, m represents the units produced by the workstation per unit time, X represents the workstation scrap percentage, {tilde over (P)} represents the workstation machining time per unit, and M represents the machine downtime;
receiving input including:
information identifying process improvement projects and corresponding predictive performance parameter values, and
information identifying resources available for process improvement;
for the at least one of the one or more workstations:
determining an achievable minimum WIP using the corresponding predictive performance parameter values, and
determining a reduction in minimum WIP based on the difference between the achievable minimum WIP for the current values of performance parameters and the predictive performance parameter values; and
determining one or more process efficiency improvements associated with the determined reductions in minimum WIP, the accounting data and the calculated standard deviation.
18. The method of claim 17 , further comprising implementing one or more of the determined process efficiency improvements.
19. A computer-readable storage device encoded with a computer program, the program comprising instructions that, when executed by one or more computers, cause the one or more computers to perform operations comprising:
generating a set of sensor data, wherein the sensor data includes measured transit time information for items flowing through a work process;
accessing a set of production control data for one or more machine tool workstations and a set of accounting data for the work process, wherein one or more of the machine tool workstations comprise (i) a cutting tool, (ii) a drilling machine, (iii) a lathe, (iv) a boring machine or (v) a grinding machine;
calculating a standard deviation of the measured transit time information;
calculating an achievable minimum work in progress (WIP) for at least one of the one or more workstations using current values of workstation performance parameters, comprising calculating an achievable minimum WIP for each workstation using
W
MIN
≅
SQm
2
(
1
-
X
-
m
(
P
~
/
(
1
-
M
)
)
)
+
Q
wherein S represents the workstation setup time, Q represents the workstation number of different part numbers, m represents the units produced by the workstation per unit time, X represents the workstation scrap percentage, {tilde over (P)} represents the workstation machining time per unit, and M represents the machine downtime;
receiving input including:
information identifying process improvement projects and corresponding predictive performance parameter values, and
information identifying resources available for process improvement;
for the at least one of the one or more workstations:
determining an achievable minimum WIP using the corresponding predictive performance parameter values, and
determining a reduction in minimum WIP based on the difference between the achievable minimum WIP for the current values of performance parameters and the predictive performance parameter values; and
determining one or more process efficiency improvements associated with the determined reductions in minimum WIP, the accounting data and the calculated standard deviation.