E3S Web of Conferences 520, 01018 (2024) https://doi.org/10.

1051/e3sconf/202452001018
ICEREE 2024

Optimization of Temperature and Humidity Monitoring for

Warehouse in the same Geographical Environment
Ronggui Dao1,a*, Yinrong Wen1, Tao Yang1, Linjia Li2,b*
1Logistics Center, Hongta Tobacco (Group) Co., LTD, Yuxi, 653100, Yunnan, China
2Logistics Management Department, China Tobacco Yunnan Industry Co., LTD, Kunming, 650000, Yunnan, China

Abstract. The purpose of this study is to solve the problem: in the warehouse temperature and humidity
monitoring, due to the wide area and large number of warehouses, a large number of monitoring systems and
equipment are arranged, which need to reduce investment and maintenance costs. Considering the same
geographical environment, select 1 or 2 typical warehouses to represent the temperature and humidity of this
region, so as to reduce the investment of monitoring equipment. We experimented with two regions separately,
with 5 warehouses in each region. Temperature and humidity were collected for a period of time, and cluster
analysis was carried out based on their maximum value, average value or comprehensive statistics of the two
regions. The results show: (1) When clustering according to the maximum temperature and humidity, there
are 2 types of significant differences warehouses in the same geographical environment. (2) When clustering
according to the mean value of temperature and humidity, there are 2 or 3 types of significant differences
warehouses in the same geographical environment, and the clustering result is different from that of clustering
according to the maximum value. (3) When clustering according to the maximum value and average value of
temperature and humidity, there is no significant difference warehouses in the same geographical environment,
which can be classified as the same warehouse. No matter considering the maximum value, average value or
the combination of the two statistics, the typical warehouse can be selected to represent the same geographical
environment, so as to reduce the monitoring equipment investment and storage management workload.

1. Introduction network is designed and implemented to monitor

environmental and agronomic variables in a plant
There are two ways to monitor the temperature and factory[1]. A digital portable low cost sensing device is
humidity of warehouse: manual monitoring or automatic designed for monitoring the sensor data of temperature
monitoring. The manual monitoring method is laborious, and humidity wirelessly with smartphone using Bluetooth
and it is difficult to ensure the timeliness of temperature technology, which monitors and analysis the data of
and humidity control. The automatic monitoring method temperature and humidity at different part of Kathmandu
applies Bluetooth[1-3], ZigBee[4-5], NB-IoT [6-9] and valley[2]. A microcontroller system is suitable to be
other Internet of Things technologies. It can be used to implemented in a software and hardware to perform
improve the accuracy of data by placing monitoring points tasks in minimizing resources, the temperature and
in places that are not accessible to humans. So it has humidity monitoring system using wireless based Xbee
become the development trend of temperature and on hydroponic plants is implemented[4]. An
humidity monitoring method. However, the warehouse environmental temperature and humidity monitoring
has a large area and a large number. Automatic system based on NB-IoT technology was developed. The
monitoring requires a large number of monitoring systems work flow of the main controller and the working mode
and equipment, and the investment and maintenance costs of the communication module were optimized to reduce
are high. the power consumption of the system[6].
In order to solve the problem of high investment in All the above studies have raised the issue of the need
warehouse temperature and humidity monitoring, many to reduce the cost in temperature and humidity monitoring.
scholars have studied it. In view of the low penetration On the whole, some studies think that the use of
rate of temperature and humidity monitoring system for automation and intelligent methods to replace manual
corn field in most farms, a set of temperature and humidity labor can reduce production costs, while ignoring the
monitoring system for corn field with high practicability system and equipment need capital investment. Some
and low cost is studied[5]. Using Intelligent Smart studies use low-cost hardware and software design, but
Farming IoT based devices, which helps in production of the accuracy is insufficient. Some studies have optimized
crops by not only upgrading it but also making it cost- the working process and the working mode of the
effective and reducing wastage[7]. A low-cost sensor communication module, but the methods are lack of

Corresponding author’s email: a*664188929@qq.com; b*28160784@qq.com

© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution
License 4.0 (https://creativecommons.org/licenses/by/4.0/).
E3S Web of Conferences 520, 01018 (2024) https://doi.org/10.1051/e3sconf/202452001018
ICEREE 2024

generality. Some studies solve the problem for specific

scenarios but lack of generalization.
Therefore, the manager considers selecting typical
warehouses to represent the temperature and humidity of
this region in the same geographical environment (There
are five main elements: warehouse building plane layout,
elevation drop similar, building structure is generally the
same, external wind direction and sunshine are basically (a)
the same). In this way, it can not only reduce the
investment, and reduce labor, but also achieve the purpose
of monitoring all warehouses in the region.

2. Materials and Methods

(b)
2.1. Materials Figure 1. Layout of temperature and humidity monitors in the
warehouse.
2.1.1. The experimental warehouse
2.2.2. Data acquisition
According to the five elements of the same geographical
environment, two regions were selected for the Every 1 hour, the monitor obtains the warehouse
experiment. The warehouses in region R1 with the same temperature and humidity, and transmits them to the
geographical environment are the 2322#, 2522#, 2722#, terminal computer through wireless transmission. Each
2922# and 3122#. The warehouses in region R2 with the monitor should acquire 779 pairs of temperature and
same geographical environment are the 4211#, 4212#, humidity data during the experiment. In R1, one monitor
4311#, 4312# and 4411#. lost packets 3 times and three monitors lost packets 1 time.
And in R2, one monitor lost packets 3 times and five
2.1.2. The monitor of temperature and humidity monitors lost packets 1 time. The monitors are working
properly (packet loss rate ≤1%).
The temperature and humidity composite monitor is used. The time series plot of temperature and humidity at
The device is based on NB-IoT, which has the each monitoring point in R1 and R2 is shown in Figure 2.
characteristics of wide coverage, massive connection, low
power consumption and low cost. The main parameters Time series plot of temperature at each monitoring point in R1
are: endurance > 730 days under 24 packets/day 20.0 Monitoring
point

transmission conditions, network delay ≤ 2 minutes, 17.5

2322A
2322B
2522A

packet loss rate ≤1%. 2522B

2722A
2722B
15.0
Temperature/℃

2922A
2922B
3122A
12.5
2.2. Methods
3122B

10.0

2.2.1. Layout of temperature and humidity monitors 7.5

in the warehouse 5.0

1 78 156 234 312 390 468 546 624 702

Two monitoring points (A, B) are set on the diagonal of Time piont

the warehouse. One of the points is set at the far end of (a)
the warehouse gate from the ground 0.3-0.5m. The other
point is set at 1-1.5 m from the top of the material. If Time series plot of humidity at each monitoring point in R1
additional points are needed, the line AB should be 70
Monitoring
point

bisected. The monitors are arranged on the aisles, or 2322A

2322B
2522A

projected horizontally onto the walls. The distance of the 60

2522B
2722A

monitoring point from the door, window and wall is not 2722B
2922A
Humidity /%

less than 1.0m to avoid radiation heat. As shown in Figure

2922B
3122A
50 3122B

1 (a), two monitors were placed in each warehouse in R1

(2322#, 2522#, 2722#, 2922# and 3122#) and R2 (4211# 40
and 4212#). And as shown in Figure 1 (b), three monitors
were placed in each warehouse in R2 (4211# and 4212#).
30

1 78 156 234 312 390 468 546 624 702

Time piont

(b)

2
E3S Web of Conferences 520, 01018 (2024) https://doi.org/10.1051/e3sconf/202452001018
ICEREE 2024

Time series plot of temperature at each monitoring point in R2

Table 1. Temperature and humidity statistics of each
18 Monitoring
warehouse in R1.
point
17 4211A
Maxim
4211B
R1 Maximum Average Average
16
4212A
um
4212B
warehou temperatur temperat humidit
humidit
4311A

se e/℃ ure/℃ y /%
15 4311B
Temperature/℃

4311C

14
4312A
4312B
y /%
4312C
13 4411A
4411B
4411C
2322# 17.4 12.6 63.0 54.5
12

11 2522# 16.4 12.9 63.9 57.5

10
2722# 17.1 13.0 68.0 61.1
9
1 78 156 234 312 390 468 546 624 702
Time piont 2922# 15.9 12.3 63.6 57.8
(c) 3122# 16.9 13.1 61.3 55.8

Time series plot of humidity at each monitoring point in R2 Table 2. Temperature and humidity statistics of each
70 Monitoring
point
warehouse in R2.
4211A
4211B
Maxim
R2 Maximum Average Average
65 4212A
4212B
um
warehou temperatur temperat humidit
4311A

humidit
60 4311B

se e/℃ ure/℃ y /%
4311C
Humidity /%

4312A

55
4312B
4312C
y /%
4411A
4411B

50
4411C
4211# 15.6 12.7 62.2 57.2
45
4212# 16.8 13.4 62.2 57.3
4311# 15.6 12.8 66.5 59.5
40

1 78 156 234 312 390 468 546 624 702 4312# 16.0 13.4 66.2 60.4
Time piont
4411# 15.7 13.0 63.5 59.5
(d)
Figure 2. Time series plot of temperature and humidity at each
monitoring point. 3. Results and Analysis

2.2.3. Data processing The validation of temperature or humidity similarity

involved multivariate analysis. Cluster analysis of
There are multiple monitoring points in the warehouse to observations was used in this study. Is it more accurate to
obtain more comprehensive temperature and humidity use the maximum value or the average value for similarity
data, so as to accurately support quality control. Which analysis? Cluster analysis is carried out on the maximum
monitoring point and what statistics are used to represent value and the average value respectively, also carried out
the temperature and humidity of the warehouse? Firstly, on the two comprehensive statistics.
given the habitual nature of statistical analysis,
temperature and humidity averages are needed. Secondly, 3.1. Analysis of warehouse similarity in R1
the maximum value of temperature and humidity is the
critical point for monitoring early warning and taking The clustering results in R1 are shown in Table 3. When
measures, which should be given special attention. The clustering by the maximum value, the temperature or
average value can eliminate the distribution difference of humidity similarity of the five warehouses shows two
temperature and humidity in the same warehouse, and significant differences, the 2322# and 3122# are one type,
only focus on the overall situation. The maximum value the 2522#, 2722# and 2922# are the other ones. When
can reflect the difference of temperature and humidity clustering according to the average value, the similarity
distribution in the warehouse. It can obtain the most also shows two significant differences, the 2322# and
unfavorable local environment in the warehouse. 2922# are one type, the 2522#, 2722# and 3122# are the
Targeted regulation should be carried out to eliminate the other ones. When clustering according to the maximum
potential hidden dangers of quality brought by local value and average value, there is no significant difference,
environment in time. One or two methods can be used in and they are all the same.
combination according to their own warehouse conditions, Summary I. In R1, if the maximum value or average
local climatic conditions, and seasonal changes, etc. So as value of the environmental parameters of the warehouse
to improve the level of storage management. is considered separately, the warehouse can be divided
The average and maximum values of temperature and into two categories, and the classification results are not
humidity of each warehouse in R1 and R2 during the the same. Considering the maximum value and average
experiment are shown in Table 1 and Table 2. value, they belong to the same warehouse, indicating that
it is feasible to select the representative warehouse to
deploy the temperature and humidity monitor for
warehouse in the same geographical environment.

3
E3S Web of Conferences 520, 01018 (2024) https://doi.org/10.1051/e3sconf/202452001018
ICEREE 2024

Table 3. Clustering results of temperature and humidity Select the representative warehouse, set the early
according to different statistics in R1. warning value, improve the monitoring frequency and
Groups Groups Groups strengthen the temperature and humidity monitoring
R1
clustered by clustered clustered by effect. Firstly, In order to facilitate centralized
warehou
maximum by average maximum and
se management, the warehouses are classified into the same
value value average value
category according to five elements: the connected layout
2322# 1 1 1
of the warehouse building plane, the similar elevation, the
2522# 2 2 1 consistency of the building structure, and the basic
2722# 2 2 1 consistency of the external wind direction and sunlight.
2922# 2 1 1 Secondly, the temperature and humidity data of each
3122# 1 2 1
warehouse of the same kind for a period of time are
counted, and the difference analysis is carried out, the
reasons for the difference are found, the preliminary
3.2. Analysis of warehouse similarity in R2 classification results are verified, and the adjustment is
made to complete the warehouse classification. Thirdly,
The clustering results in R2 are shown in Table 4. When According to the warehouse classification, select 1-2
clustering by the maximum value, the temperature or representative warehouses to represent the overall
humidity similarity of the five warehouses shows two situation of temperature and humidity in the monitoring
significant differences, the 4212# is one type, the 4211#, region.
4311#, 4312# and 4411# are the other ones. When
clustering according to the average value, the similarity
shows three significant differences, the 4211# is one type, 5. Conclusion
the 4212# is the other one, the 4311#, 4312# and 4411#
are the third ones. When clustering according to the In the same geographical environment, the typical
maximum value and average value, there is no significant warehouses are considered to be selected to monitor
difference, and they are all the same. temperature and humidity. The important statistics of
Summary II. Consistent with Summary I. temperature and humidity are the maximum and average
values. When the maximum value or average value is
Table 4. Clustering results of temperature and humidity considered separately, there are two or three kinds of
according to different statistics in R2.
significant differences between the warehouses with the
Groups Groups Groups
R2 same geographical environment, and the classifications
clustered by clustered clustered by
warehou are different. However, when the maximum value and the
maximum by average maximum and
se average value are comprehensively considered, there is no
value value average value
4211# 1 1 1 significant difference between the warehouses, which can
be classified as the same kind of warehouses. Whether the
4212# 2 2 1
maximum value, the average value or the combination of
4311# 1 3 1 both statistics are considered separately, the typical
4312# 1 3 1 warehouses can be selected to achieve the purpose of
4411# 1 3 1 reducing the investment of monitoring equipment and the
workload of warehouse management.

4. Discussion
Acknowledgments
With the development of wireless communication
technology of industrial Internet of things, the The research is supported by the following funds: 2018
temperature and humidity monitoring mode of warehouse State Tobacco Monopoly Administration standard project
will be replaced by automatic monitoring. The collected Monitoring guide for basic environmental parameters of
data is very rich, one is that the data acquisition interval tobacco aging warehouse. Project number: 2018B001.
can be shortened, and the other is that the temperature and
humidity monitor can be distributed to various parts of the References
warehouse. How to fully mine these data and give full
play to the advantages of existing technologies remains to 1. Montoya, A. P. Obando, F. A. Osorio, J. A. Morales,
be further studied？ For example, such a huge amount of J. G. Kacira, M. Design and implementation of a low-
data, only use the maximum and average value of cost sensor network to monitor environmental and
temperature and humidity to reflect the basic agronomic variables in a plant factory[J].Computers
environmental parameters in the warehouse, other series and Electronics in Agriculture, 2020, 178(1).
of data waste? In addition, how to layout the monitoring 2. Shrestha R .MONITORING AND ANALYSIS THE
points make less investment and fully monitor the DATA OF TEMPERATURE AND HUMIDITY BY
important temperature and humidity of the whole USING ARDUINO AT DIFFERENR PART OF
warehouse? So as to prevent the loss of material quality KATHMANDU VALLEY[J].Semiconductor
problems. A series of studies on data sampling frequency, Science and Information Devices, 2021.
statistics, and monitor layout were carried out[10].

4
E3S Web of Conferences 520, 01018 (2024) https://doi.org/10.1051/e3sconf/202452001018
ICEREE 2024

3. Bhanuprakash J, Reddy S V, Reddy L S.

INNOVATIVE REMOTE CONTROL SYSTEM
FOR IRRIGATION USING GSM AND
BLUETOOTH[J].[2024-01-11].
4. Emge A B, Afrianto I , Atin S. Temperature and
Humidity Monitoring System using Wireless Based
Xbee on Hydroponic Plants[J].IOP Conference
Series: Materials Science and Engineering, 2020,
879(1):012097 (7pp).
5. Weibo L, Qian H, Yue R, et al. Design of Corn Land
Temperature and Humidity Monitoring System
Based on Wireless Sensor Network[J].Journal of
Agricultural Mechanization Research, 2020.
6. Mao M , Jiaqi J U , Ouyang Y , et al. Design and
implementation of NB-IoT based environmental
temperature and humidity monitoring
system[J].Chinese Journal of Internet of Things,
2021, 5(4):99-106.
7. Rathod N, Panigrahi S, Pinjarkar V. SMART
FARMING: IOT Based Smart Sensor Agriculture
Stick for Live Temperature and Humidity
Monitoring[J].SSRN Electronic Journal, 2020(07).
8. Adhiwibowo W, Daru A F, Hirzan A M. Temperature
and Humidity Monitoring Using DHT22 Sensor and
Cayenne API[J].Jurnal Transformatika, 2020, 17(2).
9. Anggraeni H E, Setiawan A, Irawan S. Temperature
and Humidity Monitoring System Environmental Cat
Incubator Based on the Internet of Things
(IoT)[J].ICVEAST, 2023.
10. YC/T598-2023 Monitoring guide for basic
environmental parameters of tobacco aging
warehouse [S].