2017 Devices for Integrated Circuit (DevIC), 23-24 March, 2017, Kalyani, India 307

Data Centre Temperature Monitoring with ESP8266

Based Wireless Sensor Network and Cloud Based
Dashboard with Real Time Alert System

Saraswati Saha Anupam Majumdar

Dept. of Electronics and Communication Engineering, Former Head of State e-Governance Mission Team (SeMT),
RCC Institute of Information Technology Tripura
Kolkata, India National Institute for Smart Government (NISG), India
[email protected] [email protected]

Abstract— The Internet of Things (IoT) system proposed in this Sensor based Temperature monitoring system is an important
paper is an advanced solution for monitoring the temperature at part of the proposed framework for an optimal environmental
different points of location in a data centre, making this monitoring strategy. The sensors should be placed at multiple
temperature data visible over internet through cloud based points of individual racks to measure the heat being generated.
dashboard and sending SMS and email alerts to predefined
recipients when temperature rises above the safe operating zone
The sensors can also be put at the air conditioning system’s
and reaches certain high values. This helps the datacenter vents, to measure efficiency. Much of the work performed so
management team to take immediate action to rectify this far for monitoring the data center environment is based on
temperature deviation. Also this can be monitored from Arduino hardware platform for data processing which is again
anywhere anytime over online dashboard by the senior level networked with XBee RF modules to develop the wireless
professionals who are not present in the data centre at any point sensor network.[1] It’s found to be an expensive solution even
in time. This Wireless Sensor Network (WSN) based monitoring with the basic XBee modules considered for the nodes. On top
system consists of temperature sensors, ESP8266 and Wi-Fi of this as these modules need some additional board like
router. ESP8266 is a low power, highly integrated Wi-Fi solution Arduino, for processing; the Total Cost of Ownership (TCO) of
from Espressif. The ESP8266 here, in this prototype, connects to
‘Ubidots’ cloud through its API for posting temperature data to
this solution rises even more. The data rate (250 Kbps) is not at
the cloud dashboard on real time and the cloud event all impressive as well. A different approach is also observed in
management system generates alerts whenever the high previously done works and studies deploying Arduino and
temperature alert event is fired. Cloud events need to be ESP8266. In this approach entire processing is done within
configured for different alerts beforehand through the user Arduino and as the WI-FI provider ESP8266 is used. Here the
friendly user interface of the platform. It’s to be noted that the cost has come down a lot compared to the first approach. But
sensor used here can be leveraged to monitor the relative the capabilities of ESP8266 are not well utilized; specifically
humidity of the data center environment as well along with the the processing power of ESP8266 is not at all used. Its usage is
temperature of the data center. But for this prototype solution kept limited for WI-FI purposes only. Now the solution being
focus is kept entirely on the temperature monitoring.
proposed here uses ESP8266 which has processing capability
Keywords—IoT; Internet of Things; WSN; Wireless Sensor as well as integrated Wi-Fi on the chip itself. No additional
Network; Data Centre; Sensor; Cloud; ESP8266; Data Center; board like Arduino is used in the proposed approach. ESP8266
Arduino IDE; cloud based event management system; DHT 11; does the processing as well as the WI-FI activities but with
DHT 22; Real time dashboard; Environmental monitoring pretty low power consumption. Data fetched by the sensors are
sent to the ESP8266 which then can be analyzed and processed
I. INTRODUCTION by the ESP8266 on board microcontroller itself with the help of
the firmware loaded into its flash memory. Wi-Fi Router acts
Data center environmental control is a constructive generic
as the Access Point of the Wireless Sensor Network or WSN.
framework for maintaining temperature, humidity, and other
On the chip Wi-Fi of ESP8266 connects to the Wi-Fi router as
physical qualities of air within a specific range in order to
a station node and can push the live sensor data to the Ubidots
allow the equipment housed in a data center to perform
cloud platform in the internet at regular intervals. That means
optimally throughout its lifespan. Here an IoT approach has
the ESP is playing as the sensor interface unit in one hand and
been taken towards building such a framework. The term IoT is
on the other hand is playing as the Wireless Communication
broadly used to describe a framework where sensors are
Unit besides doing all the processing. This results into faster
connected to objects and help these objects to share their
processing and transmission. This solution lowers the Total
‘digital voice’ with the external world over internet connection.
Cost of Ownership (TCO) as well compared to the two
978-1-5090-4724-6/17/$31.00 ©2017 IEEE
 2017 Devices for Integrated Circuit (DevIC), 23-24 March, 2017, Kalyani, India 308

approaches discussed earlier. Current trend of IoT focuses on

pushing data into a centralized cloud platform [2] which is
B. Hardware Implementation
being administered by IT experts and available to anyone with
the proper credentials and an Internet connection. The power The hardware components used in the proposed system
of cloud computing is making the required storage and need some discussion.
processing power available to handle huge sensor data. The Measurement range of DHT 11 is from 0 degree centigrade
overall uptime of these platforms continues to trend upwards to 50 degree centigrade with response time of 6 to 30 seconds.
as they become more resilient to the increasing demand and Its small size, low power consumption and maximum 20 meter
expectations of the IoT world. [3][4] signal transmission makes it the suitable for this application.
Being aligned with this trend, the proposed approach also For wider temperature range and better precision DHT22
includes Cloud implementation on top of ESP8266 WSN based sensors are used.
temperature monitoring system. Here not only the dashboard
feature of the cloud solution is leveraged but also the ‘Alert / ESP8266 is a system-on-a-chip (SoC) designed by Espressif
Feedback system from cloud’ is implemented and tested. Systems which is based on 32-bit RISC CPU with the Tensilica
Xtensa LX106 processor. It has features like inbuilt Wi-Fi
(802.11 b/g/n), GPIO (General Purpose Input / Output), Inter-
II. SYSTEM OVERVIEW Integrated Circuit (I²C), analog-to-digital conversion, Serial
Peripheral Interface (SPI), UART (Universal asynchronous
A. System Components receiver/transmitter), and pulse-width modulation (PWM).
DHT 11 sensors send the temperature data to the ESP8266. ESP8266 has a number of versions with varying capacities.
For every single test location, 2 sensors are placed to validate
For this project its basic level version i.e. ESP8266-01 will
received sensor data as well as to provide device redundancy at
be sufficient enough with its 2 GPIO pins and integrated
the sensor level. ESP8266 is loaded with the firmware program
TCP/IP protocol stack. [5] In this project ESP is used as a
written in C that does all the interfacing with sensors,
station and it connects to the Access Point of the wireless
processing the sensor data and interfacing with cloud platform
network i.e. the wireless router using the SSID (Service Set
and finally uploading the data to the cloud platform ideally
Identifier) and the password of the network. As it connects to
once every one minute. To write the code and upload it to the
the Access Point successfully it gets assigned to an IP.
ESP8266, Arduino IDE is used. Two sets of data from two
sensor units are compared for an unusual mismatch and if any Now to get the desired functionality out of the ESP, it needs
such mismatch is found that set of data is discarded and next to be programmed. This is done here by connecting the ESP to
set of data are fetched. This processing is done by the Arduino IDE (Integrated Development Environment) running
microcontroller along with supporting components on the on a laptop through the UART interface (Transmit and Receive
ESP8266 board. If data received from two sensors located at a pins) of ESP via UART to USB converter (figure 2). Compiled
single spot are found somewhat matching, then that set of data binary version of the C program is uploaded to the flash
is pushed to the Ubidots cloud. Ubidots cloud platform needs memory of the ESP.
to be configured to accept this data from the sensors and show
the data over dashboards. This platform also has feature to
configure alert for different events. For this case, multilevel
alert SMSs are configured for specific phone numbers.
Similarly email notifications are configured to report alert at
multiple levels of temperature.

Figure 2: Uploading of firmware to ESP8266

Figure 1: System Diagram

 2017 Devices for Integrated Circuit (DevIC), 23-24 March, 2017, Kalyani, India 309

C. Software Implementation
• Arduino IDE (Integrated Development
Environment) – This IDE can be used to program III. FLOW STEPS
ESP8266 as well.[7] The firmware is written in C
language including required header files to • Sense the temperature at a specific location of a
integrate DHT sensors. Also the program includes rack with the help of 2 sensors.
header file required to interact with the cloud • ESP8266 receives the data from the sensors
platform. This IDE helps to write and compile the through its GPIOs (General Purpose Input Output)
firmware and then upload to the flash memory of
• If the values received from 2 sources are not
the ESP8266. The library files required for the
diverse beyond a certain limit, then the values are
prototype are extracted in the folder pointed as
considered to be posted, otherwise that set of data
library folder for the Arduino IDE. [8]
are discarded by the ESP8266.
• Cloud Platform: The ESP8266 and sensors need to
• ESP8266 connects to the Wi-Fi router using SSID
be integrated first with the cloud platform.[9]
and password.
Ubidots cloud platform provides user friendly
interfaces to do this and assigns a token for every • ESP8266 also establishes connection to the cloud
device or sensor variable. Once configured this platform using http protocol.
cloud platform provides online real time
dashboards for the sensor temperatures. The same • ESP posts temperature data to the cloud platform
platform also provides interface to define SMS using the token assigned. [9]
and email alerts on specific sensor values. • The cloud platform shows the live real time data
on the dashboard.
• This cloud platform shoots alert SMS and alert
emails as and when predefined alert situation
arises i.e. when temperature rises beyond alarm
settings.

Figure 3: Cloud configuration for sensors

Figure 5: Prototype Setup at lab with DHT 11 and ESP8266

IV. RESULTS

As the prototype system is deployed at a setup, the live data

is being monitored over the cloud dashboard. The cooling
system is adjusted a bit to raise the temperature beyond a set
limit for a while. The cloud alert system sends out SMS alert to
preconfigured cell phone as the temperature touches the alarm
setting value. It also sends out a mail to the defined email id.
The dashboard continuously presents the live data. As the
Figure 4: Cloud configuration for Alerts cooling system is made normal, the temperature gradually
corrects towards normal one. This is also depicted over the
 2017 Devices for Integrated Circuit (DevIC), 23-24 March, 2017, Kalyani, India 310

dashboard. Dashboard snippets for uprising and falling References

temperature are shown here.
[1] Michael G. Rodriguez, Luis E. Ortiz Uriarte, Yi Jia, Kazutomo Yoshii,
Robert Ross, Peter H. Beckman “Wireless Sensor Network for Data-
Center Environmental Monitoring,” published in 2011 Fifth
International Conference on Sensing Technology (ICST)
[2] Industrial Internet of Things and Communications at the Edge by Tony
Paine
[3] Designing the Internet of Things by Adrian McEwen and H. Cassimally
[4] The Internet of Things in the Cloud by H. Zhou
[5] http://espressif.com/ ESP8266 Details
[6] http://www.esp8266.com/ Information on ESP8266
[7] https://www.arduino.cc Arduino software download
[8] http://playground.arduino.cc/Main/DHT11Lib DHT library
[9] https://ubidots.com/ Ubidots Cloud Configuration
[10] http://download.arduino.org/products/UNOWIFI/0A-ESP8266-
Datasheet-EN-v4.3.pdf Datasheet for ESP8266

Figure 6: Temperature Rise as seen on dashboard

Figure 7: Temperature Fall towards Normal

V. CONCLUSIONS
A prototype wireless sensor network for data center
temperature monitoring is developed with the help of
ESP8266. A continuous test run of more than 24 hours shows
steady performance of ESP8266. This system not only adds up
value to the existing on site monitoring system available at data
centers but also provides lot of comfort to the higher authority
of Data Center management to monitor the parameters from
remote anytime with the help of the dashboard. The SMS and
email based alert system adds additional layer of tighter
monitoring of the vital parameters thereby assisting to take
immediate action if any alarming situation is reported.
ESP8266 with its very compact size, higher data rate, low
power operation and very low TCO makes the solution more
deployment ready for real life scenarios.