Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 196 (2017) 121 – 127

Creative Construction Conference 2017, CCC 2017, 19-22 June 2017, Primosten, Croatia

A Prediction Method of Vertical Lifting Loads based on Long-Term

Time Series Analysis for High-Rise Building Construction
Kyuhyup Leea, Soonwook Kwon*, Joonghwan Shinb, Suwan Chungc and Songyi Leed
a
Department of Convergence Engineering for Future, Sungkyunkwan University, 2066, Seoburo, Jangan-gu, Suwon-si, Gyeonggi-do, Republic of
Korea
*
School of Civil & Architectural Engineering, Sungkyunkwan University, 2066, Seoburo, Jangan-gu, Suwon-si, Gyeonggi-do, Republic of Korea
b
Department of Convergence Engineering for Future, Sungkyunkwan University, 2066, Seoburo, Jangan-gu, Suwon-si, Gyeonggi-do, Republic of
Korea
c
Department of Convergence Engineering for Future, Sungkyunkwan University, 2066, Seoburo, Jangan-gu, Suwon-si, Gyeonggi-do, Republic of
Korea
d
Department of Convergence Engineering for Future, Sungkyunkwan University, 2066, Seoburo, Jangan-gu, Suwon-si, Gyeonggi-do, Republic of
Korea

Abstract

As recent building have been getting higher and larger, vertical lift planning and operation are key factor for successful project.
Although many studies have been trying to set up a construction lift planning system at early stage before lift installation, but there
are not regarding lift management for on-going project using operation data from the field. In this study In this study, to predict
lifting event which may be generated during remained construction schedule, long-term time-series based data analysis was
conducted. Data was collected embedded lift monitoring sensor and database. The pattern analysis was carried out reflecting
characteristics of Tact based finishing work control. And this study shows an effective value of activity oriented time-series analysis
to lift management on high-rise building construction. Lastly, this study has served as a key lead of lift monitoring based decision-
making for vertical lift zoning. Therefore, it is anticipated to settle as a system based data-centric construction management
technique in the field.
©©2017
2017TheThe Authors.
Authors. Published
Published by Elsevier
by Elsevier Ltd.is an open access article under the CC BY-NC-ND license
Ltd. This
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017.
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017
Keywords: Construction Lift, Finishing work schedule, Time-series analysis

* Corresponding author. Tel.: +82-31-299-4765; fax:+82-31-290-7570.

E-mail address: swkwon@skku.edu

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2017
doi:10.1016/j.proeng.2017.07.181
122 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127

1. Introduction

In high-rise building construction, finishing work comprises approximately 15–20% of the total construction cost.
The most critical problem in the finishing work phase in high-rise building construction is lift management in terms
of the manpower and materials required. The main reason for lift management problems is the long distance required
in vertical transportation and the increase in materials required for finishing work.
In the Republic of Korea, high-rise buildings over 500 m generate an additional cost of over $20 million due to
finishing work delays of over 3 months. These additional costs are incurred very frequently, which demonstrates
problems in the current lifting plans used in most high-rise construction projects.
The existing method for lift management, including the lift load for each construction phase and the peak time
period, is based on a simple method that is generally used in buildings under 300 m. However, a method to correct the
existing errors in lift management is required for high-rise building construction.

2. Research trends

Sacks et al. proposed an automated lift equipment monitoring system [1], and Cho et al. carried out a study on
construction lift operation planning in terms of lifting height and loading [2]. In terms of metaheuristic approaches,
Shin proposed the optimal operation of temporary construction lifts in super high-rise building construction based on
a simulation and genetic algorithm [3]. Before these studies, most research focused on the use of tower or mobile
cranes, and studies on construction lifts tended to emphasize lift planning rather than lift operation.
As super high-rise construction becomes more popular in Korea, there is a growing need for systematic construction
planning and site management. Government and private corporations are actively undertaking studies on the
operational planning of construction lifts and tower cranes. For example, Kim studied how to calculate the number of
construction lifts necessary at super high-rise construction sites [4], and Cho proposed an algorithm to calculate the
lifting time according to the acceleration and deceleration capabilities of construction lifts [2]. While many studies
have been conducted on construction lift planning, few data-based methods have been introduced on the system,
management, and algorithm of construction lift operation.

3. Operational lift management

3.1. Problems of plan-based lift management

Both temporary lift-based finishing materials and worker management have a direct impact on construction delays
and cost increases due to the nature of high-rise projects having long vertical distances of over 500 m. With lift
equipment that lifts 100 m/min, it takes 4 minutes to reach the top of a 400 m mast (i.e., the maximum height that can
be reached according to the current technology level). The operating time is increased by 6% per stop at each floor,
with 15 seconds as the average stopping time per story. In the construction of high-rise buildings, the efficiency
degradation of lift operation increases exponentially.
Solving the problems caused by the characteristics of vertical lifting in high-rise construction projects cannot be
expected in the short term. At present, lift-based vertical transportation planning and construction for high-rise projects
are established using a heuristic model based on the following criteria. The current heuristic equations used for high-
rise construction are as follows in equations (1) and (2).

Number of lifting Material quantities u Total floor area (1)

Total Labor
Number of lifting (2)
Boarding Capacity
 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127 123

Fig. 1. Lift planning process

In lift planning, as shown in Fig. 1, the criteria and weighting factor used for lift calculations like those in (1) and
(2) are based on data from Japanese construction companies. However, no studies of their suitability for use in other
high-rise construction sites in other countries have been conducted. This is a major limitation of plan-based lift
management, as shown in Fig. 2.

Fig. 2. Difference between predicted and actual labor lifting

3.2. Lift management based on the operational phase

Even though lift management should be planned prior to construction, there are two reasons for a construction
manager to focus on operation instead.
First, the decisions surrounding lift loads and selection in the planning phase are difficult to use as reference data
for all high-rise building projects. However, lift operation data analyzed in the construction phase can be utilized after
the implementation of the project, as the operation data reflects the actual statement of the building according to the
construction process.
Second, in dynamic construction environments, operational monitoring data provides more practical support to
decision making regarding the lift control suitable for the current state than lift planning does. In addition, differences
between planning and operation in terms of the definition of a single lifting event’s cycle time exemplify the
uncertainty of the planning equation in terms of the expected operation result and the actual result, as shown in
equations (3) and (4).

Hu2
T  ( ML u F ) (3)
VuA

Equation (3) is currently used for lifting plans. In equation (3), T indicates the total moving time for a single cycle,
H indicates the maximum height the lift can achieve, V indicates the moving speed of the lift, ML indicates the time
spent loading and unloading materials, and F indicates the number of stops. The uncertainty of equation (3) is generated
by variable A, as variable A indicates the operational efficiency, and no practical basis exists when managers apply
variable A as the operational efficiency.

n
H n  h1  h2
T ¦{t
k 1
1 (
V
)  TDn } (4)
124 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127

Equation (4) is used to calculate the operational efficiency of a construction phase. In equation (4), tଵ indicates the
acceleration time, H୬ indicates the lifting height of n୲୦ lifting events, t ଶ indicates the deceleration time, hଵ indicates
the distance for acceleration, and hଶ indicates the distance for deceleration. In this equation, TD୬ indicates the time
required for the dispatch demands of additional hardware to measure workers’ waiting time. TD୬ shows definite
evidence for lift efficiency.

4. Vertical zoning of construction lift operation

4.1. Characteristics of vertical construction lift zoning

The higher the building, the longer finishing work is delayed. In addition, most high-rise buildings are built as set-
back buildings due to issues in their slenderness ratios. Therefore, the operation zones, including the floor division and
set of transfer areas, are critical to consider in order to achieve the vertical lifting of workers and materials.

Table 1. Characteristics and differences between elevators and construction lifts

Elevator Construction Lift

Purpose To transport residents to each floor To lift materials and workers to each work site
Traffic Calculation a) Transportation capacity per 5 minutes a) Lift load for finishing work material
b) Occupied area per man b) Lift load for worker (operation frequency)
Moving Section Fixed (i.e., the top floor of the building is Changeable (i.e., the top floor is changeable
never changed) according to the schedule)
Zoning Criterion To reduce the ratio of the core span area to the To minimize the frequency of vertical lift
lease span area operation

As shown in Table 1, in terms of the purpose of use and traffic calculation, the characteristics of building elevators
and construction lifts are different, though they are functionally and technically similar. Generally, building elevator’s
moving sections are decided not only by traffic control but by the use of each floor. The vertical zoning of high-rise
construction lifts aims to minimize the frequency of vertical lift operation and to reduce workers’ waiting time. A
changeable moving section according to the construction schedule is included after zoning decisions are made.

Fig. 3. Vertical zoning of construction lift operations

 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127 125

4.2. Characteristics of the finishing work process

Construction lift operation is deeply related to the finishing work process and its control techniques. It is important
to determine a reliable pattern of lift movement by reviewing accumulated history. However, it is not easy to handle
dynamic pattern changes before or after finishing work activity.
To avoid interference and conflicts in each activity in order to better control finishing work schedules, some process
techniques are used, such as the line of balance and TACT methods. The purpose of both techniques is to standardize
the schedule of activities under a linear pattern.

Fig. 4. Example of finishing work activity management

5. Vertical zoning of construction lift operation

5.1. Embedded lift operation monitoring

To analyze lifts’ patterns of operation, collecting their movement history in real time is essential. Fig 5 represents
the embedded data box developed to collect movement data. This embedded lift monitoring system receives the
location, speed, load, and direction from sensors (i.e., an encoder, a load cell, a converter, a current sensor, a limit
switch, and a proximity sensor).

Fig. 5. Data box developed for operational analysis

In order to perform environmental data analysis, the embedded lift monitoring system generates every lifting event
along with its signal time, target floor, and lifting weight, as shown in Table 2. The variable t indicates signal time of
each lifting event, E indicates the events, f indicates the floors, and w indicates the lifting weight.
126 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127

Table 2. Data box arrangements for predicted lifting events

Single Time (t) Target Floor (f) Lifting Weight (w)
Lifting Event 1 (t ଵ , Eଵ ) (fଵ , Eଵ ) (wଵ , Eଵ)
Lifting Event 2 (t ଶ , Eଶ ) (fଶ , Eଶ ) (wଶ , Eଶ )
Lifting Event 3 (t ଷ , Eଷ ) (fଷ , Eଷ ) (wଷ , Eଷ )
Lifting Event 4 (t ସ , Eସ ) (fସ , Eସ ) (wସ , Eସ )
Ή Ή Ή Ή
Lifting Event n (t ୬ , E୬ ) (f୬ , E୬ ) (w୬ , n)

5.2. Relationship between time-series analysis and finishing work

In the construction field, the floor area is different in each floor section. In contrast, equations use unclassified floor
area values to calculate the total lifting load. Still, the accumulated operation data, including the lifting weight and the
target floor, show a direct correlation with delays in finishing work.
The dataset collected by monitoring sensors seems to be disordered. However, the large data stream based on
timetables provides predictable patterns for each finishing work activity, each of which follows certain standards. In
other words, the arrangement of lifting load data shows a similar shape with finishing work schedules according to the
line of balance and TACT techniques. When temporary construction delays emerge, managers are able to control data
systems by checking the date of the delay.
In activity-oriented lifting analysis, another important issue is the division of sections, as shown in Table 3. In a
private office in a building in Jamsil, Korea, for example, the area of the room is presented as 1281.20 × 0.95337୲ .
Even though most high-rise buildings use floors in diverse ways according to the height of the floor, existing lift
planning methods apply the area values of typical floors without variation.

Table 3. Data box arrangements for predicted lifting events

Floor Floor Section
Podium B1~12F
Primary Office 15F~39F
Apartment 44F~72F
Hotel 76F~101F
Private Office 106F~114F
Observation Deck 117F~123F

5.3. Activity-oriented, long-term time-series analysis of operation

The primary office (15~39F) floor areas do not change according to floor height. In the case of tiling ceilings, the
prediction value of floor 39 from the cycle-based lifting load data along 15~38F is 7.932 tons, as shown in Fig. 6. This
predicted value is similar to actual value of 8.321 tons.
 Kyuhyup Lee et al. / Procedia Engineering 196 (2017) 121 – 127 127

However, the apartment section shows an exponential decrease. In the data on ceiling m-bar work, the exponential
equation of the lifting load is 11.9446 × 0.985525୲ . The prediction value of floor 72 from the cycle-based lifting load
data along 44~71F is 7.839 tons, as shown in Fig 7, which is similar to the actual value of 6.909 tons.

Fig. 6. Predicted lifting load using activity-oriented time-series analysis:

(a) tiling ceiling work in the primary office and (b) ceiling m-bar work in the apartment section

6. Conclusion

In high-rise buildings, limits on the maximum lifted floors increase, and the variables that must be considered by
management when finalizing construction become complicated. Thus, the operational management of the lift becomes
a more important consideration than lift planning. The results of operational management can be used in construction
along with a decision-making procedure on the transportation and circulation of finishing work materials, both of
which are altered to fit the situation at hand.
In this study, to predict lifting events that may occur during construction, data analysis based on a long-term time
series was conducted. The data was collected through embedded lift monitoring sensors and databases. A pattern
analysis was conducted to examine the characteristics of TACT-based finishing work. This study shows the
effectiveness of activity-oriented time-series analysis in lift management for high-rise building construction. In
addition, this study provides valuable information on decision making in terms of vertical lift zoning and monitoring
as well as a systematic and data-centric construction management techniques for practical applications.

Acknowledgements

This research was supported by a grant(15AUDP-B106327-01) from architecture & Urban Development Research
Program funded by Ministry of Land, Infrastructure and Transport of Korean government.
This work is financially supported by Korea Ministry of Land, Infrastructure and Transport(MOLIT) as ೻U-City
Master and Doctor Course Grant Program೼.

References

[1] R. Sacks, R. Navon, I. Brodetskaia and A. Shapira, Feasibility of Automated Monitoring of Lifting Equipment in Support of Project Control, J.
Constr. Eng. Manage, 2005, pp. 604-614
[2] C.Y. Cho, Y.S. Shin, S.K. Kim, J.Y. Kim and M.Y. Cho, An Algorithm for Hoisting Time Calculation in Super-tall Building Construction,
Korean journal of Construction Engineering and Management, 2011, pp. 120-129
[3] S.K. Kim, K.K. Han, A Study on the Estimation of Proper Number of Construction Lifts, The Korea Institute of Building Construction, 2008,
pp. 119-125
[4] Y.S. Shin, Construction Lift Planning System in Tall Building Construction, Architectural Institute Korea, 2011, pp. 227-235