In this article, we examine a correlation between environmental information such as temperature and humidity and the concentration of desk job. We obtained environmental information from OMRON's Environment Sensor and obtained concentration information from a device that measures brain waves. We compared both by a machine-learning tool, and we found that there was a correlation between environmental information and concentration of desk jobs. We also propose that concentration could be estimated simply by using Environment Sensor.
IntroductionThese days all the employers targets to improve the efficiency of desk jobs. Some companies even try to monitor the concentration of each employee constantly in order to increase the productivity at work. The degree of concentration can be measured by brain waves, but it is difficult to always wear a device that measures brain waves. In order to verify if concentration can be measured without contacting any device to head, we examined the possibility of estimating the degree of concentration from the environmental information such as temperature and humidity.
We used OMRON environment sensors in order to get environmental data. The Omron environment sensor can acquire various environmental information of a designated area in real time. The sensor outputs parameters such as temperature, humidity, light, UV, pressure, sound, VOC, acceleration, etc. This sensor can communicate with devices (smartphones, tablets, gateways, etc.) via Bluetooth Low Energy and USB2.0 communication interface. It is attractive because the quality is guaranteed, as it is a final product.
We used following things for our evaluation.
- Omron Environment Sensor 2JCIE-BU01
https://www.components.omron.com/product-detail?partId=73065
- Cassia Bluetooth Router E1000
https://www.cassianetworks.com/products/e1000-bluetooth-edge-router/
- Device to get brain waves
Omron Environment sensor has USB and Bluetooth communication protocols and can be connected to various gateways. In this article, considering the sensor to be used in large places such as an office; we used Cassia Bluetooth Router E1000 for our system. It extends the Bluetooth range up to 1000 feet in case of open space without requiring any change to sensor.
We made an environmental data acquisition system using Cassia's router.
Omron's Environment sensor data is stored in Amazon S3 via Cassia router E1000.
Omron Environment Sensor 2JCIE-BU01 technical information:
Link for user manual:
https://omronfs.omron.com/en_US/ecb/products/pdf/A279-E1-01.pdf
Link for sample code for Raspberry Pi (USB communication):
https://github.com/omron-devhub/2jciebu-usb-raspberrypi
Link for sample code for Raspberry Pi (Bluetooth communication):
https://github.com/omron-devhub/2jciebl-bu-ble-raspberrypi
Cassia Router E1000 :
Link for user manual:
https://www.cassianetworks.com/download/docs/Cassia-User_Manual.pdf
Link for SDK: https://www.cassianetworks.com/download/docs/Cassia_SDK_Implementation_Guide.pdf
https://github.com/CassiaNetworks/CassiaSDKGuide/wiki
Link for knowledge base:
https://www.cassianetworks.com/support/knowledge-base/
3: Get data from environment sensor:In order to perform the verification, we connected the environment sensor to PC's USB port to power it.
The subjects attached brain wave measurement devices to their heads to monitor brain waves.
While conducting normal desk jobs in the office, we acquired both brain waves information and environmental information.
Environment sensor information was stored on the AWS cloud. We used the data for temperature, humidity, light, barometric pressure, sound noise, eTVOC and eCO2 from cloud. We performed data processing before analyzing the data. We filled in the values for the data loss and implemented time parameter processing algorithm.
The brain wave data obtained from the measurement device is added with the environmental data to create one CSV file containing all the data. For this article, we used β waves suitable for simple tasks and θ waves that shows creativity, and analyzed the correlation between them and the environmental information.
We performed data analysis using Amazon SageMaker as machine learning tool. We verified if there is a correlation between environmental information and brain waves that indicates the degree of concentration.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from IPython.display import Image
from IPython.display import display
from sklearn.datasets import dump_svmlight_file
from time import gmtime, strftime
import sys
import math
import json
import os
import sagemaker
from sagemaker.predictor import csv_serializer
data = pd.read_csv('./20191205221560_11to17-cassia_data_converted_merge_13.csv')
display(data.corr())
pd.plotting.scatter_matrix(data, figsize=(12, 12))
plt.show()
Following is the result we obtained.
If the correlation value is 0.3 or higher, it can be determined that, there is a correlation between both the data’s.
Our analysis showed that Temperature, Humidity and eCo2 are correlated with β and θ waves.
We found that if we use a pre-trained machine learning model engine, we could easily estimate concentration using only environmental sensors.
Conclusion:We obtained environmental information from OMRON's Environment Sensor and obtained concentration information from the device that measures the brain waves. We compared both with a machine-learning tool, and found that there is a correlation between environmental information and concentration information for desk jobs. In addition, we found that concentration could be estimated simply by using Omron’s Environment Sensor.
Comments