This post is also available in: Deutsch (German)
This article looks at why it makes sense to use sensor data analytics in public transport.
Sensor data analytics
Sensor data analytics is said to analyse the data collected from sensors and IoT devices to measure the required metrics of a device either in real-time or at specific time intervals.
What is IoT?
IoT, which stands for the Internet of Things, is the network of web-enabled smart devices that use sensors to collect data from their surroundings, process the sensor data and then transfer the sensor data via an IoT gateway that leads data to the cloud or a network protocol leading to a data reservoir.
Why is IoT sensor data beneficial?
The automatically collected, processed and transferred sensor data allows real-time monitoring and decision-making, requires minimal human intervention (which is cost-efficient) and can be accessed anywhere and anytime.
According to the environmental engagement platform Temboo, sensor data can be used in 5 ways:
- “Live Data: I want to know when something is not working.
- Historical Data: I want to keep logs of when something has and has not been working.
- Analytical Data: I want to understand why something isn’t working.
- Predictive Data: I want to know when something will stop working.
- Data for Change: I want to change how something works.“
Whichever way a set of sensor data is used depends on the purpose of using the sensor data set and the types of sensors used to collect the data. As an example, the Android Developers’ guide to sensors seems to indicate that there are 3 types of built-in sensors for different purposes (at least, in Android devices):
- Motion sensors to detect movements such as accelerations and gravitational forces.
- Environmental sensors to measure the characteristics of an environment such as temperature, pressure, humidity and light.
- Position sensors to locate physical positions.
Conversely, different types of sensors can be used to collect different parameters for the same purpose. For example, motion sensors, seat pressure sensors, vehicle weight sensors and heat sensors are different options of people counting methods using sensor data in public transport.
However, the availability of sensor data thanks to the installation of sensors will be nothing without the right system in place to take advantage of the sensor data. Sensors of Smart Devices in the Internet of Everything (IoE) Era: Big Opportunities and Massive Doubts by Mohammad Masoud et al (2019) made a case for identifying the right methods like event-driven programming for processing and deriving meaningful insights from sensor data.
The combination of IoT devices with automated systems for gathering data, analysing the data and making decisions based on the data can be seen in use cases like water management, garbage collection, smart cities, restaurant management and cycling navigation.
Maybe that’s why the IoT technology consists of 3 levels:
- IoT devices for collecting data
- IoT gateway for transfering data
- IoT platform for using data
Sensor data use cases for public transport
Cities Today quoted the expensive Smart City Use Cases & Technology Adoption Report 2020 report, which said that connected public transport has taken the top spot for IoT use cases with an implementation rate of 74% in early 2020.
To see how this is possible, perhaps we can zoom in closer to some of the sensor data use cases for public transport below.
1. People counting for occupancy data and ticket revenue sharing
Counting the number of people in public transport is both crucial for monitoring occupancy or crowding levels and fairly distributing the revenue of transport tickets. The examples below prove that sensor data can be used to count people.
East Denmark’s public transport authority, Movia, installed IoT sensors in its buses as part of its pilot project to count the number of smartphones with internet signals to gauge the occupancy levels of buses.
Similarly, Ryu, Park and El-Tawab presented the Wi-Fi sensing system as a low-cost transport data collection system in their 2020 paper WiFi Sensing System for Monitoring Public Transportation Ridership: A Case Study. Testing in 4 bus stops in Charlottesville, Virginia, USA, the team found that the Wi-Fi sensing system backed by the sliding window algorithm can estimate each bus passenger’s journey, waiting times and ridership with minimal errors.
2. Transport planning and improvements
Between January 2017 and September 2018, a German consortium research project, Digitale Mobilität – Fahrzeug und Haltestelle (DiMo-FuH), tested the application of the IoT protocol MQ Telemetry Transport (MQTT) in facilitating standardised and reliable communications via info systems like displays, audio and ticket machines in stations and vehicles.
Also in 2017, Rinne, Bagheri and Tolvanen studied the Automatic Recognition of Public Transport Trips from Mobile Device Sensor Data and Transport Infrastructure Information. To understand the transport habits of passengers in Helsinki to inform transport route planning, the trio compiled automatically-tracked locations of passengers’ mobile device sensors, which were then cross-checked with daily manual logs of public transport trips.
3. Real-time operation management
Making sure that public transport runs smoothly is an ongoing process, thus requiring real-time sensor data to drive or even automate daily operational decisions. Here’s one crucial example of using real-time sensor data in operation management.
LoRaWaN hardware provider Semtech and Montreal-based LoRaWaN network operator X-TELIA joined forces in 2018 to install LoRa-based solar-powered display screens at bus stops to show accurate real-time bus schedules for passengers who want to know when the next bus is coming.
LoraWaN is characterised by its variations in long-range network architecture, low power consumption, network capacity, network security, robustness to interference, bidirectional communication and potential IoT applications. Such characterisation can be seen in its device classes A, B and C, which reflect the trade-off between downlink network communication latency and battery lifetime, with class A leaning towards the former and class C leaning towards the latter. Put simply, latency eats up battery.
It was also for these reasons that telecom operator Tele2 and IoT network specialist TalkPool joined forces to deploy LoRaWaN in the Swedish city of Gothenburg for its smart city project in 2016.
In another example, Slovenian intercity transport provider Nomago brought together 12 different bus operations across Central and Eastern Europe into 1 transport management platform. To help with the streamlining, IoT devices such as GNSS, IMU and CAN BUS were installed in the city and regional buses to gather the sensor data into the transport management platform.
Similarly, He, Hu, Park and Levin delivered an overview of the Vehicle sensor data (VSD)-based transportation research: Modeling, analysis, and management. The team split VSD data into 2 categories: location-based VSD (LB-VSD) and surrounding traffic VSD (ST-VSD).
LB-VSD can be derived from passengers’ mobile device sensors and vehicles’ Global Positioning System (GPS) sensors. Being location-oriented, LB-VSD can be used for following routes, measuring travel times, forecasting traffic conditions and coordinating e-hailing services.
ST-VSD can be derived from onboard sensor tech such as Light Detection And Ranging (LIDAR), Controller Area Network (CAN) bus, Dedicated Short Range Communication (DSRC) to monitor and manage surrounding traffic conditions in a Connected and Automated Vehicles (CAV) or driverless environment.
So, how can sensor data be the secret ingredient for public transport?
BLE, Wi-Fi sensing, MQTT, mobile sensors, LoRaWaN, GNSS, IMU, CAN BUS, GPS sensors, LIDAR and DSRC. These are the various types of IoT sensor tech that can be deployed in the transport setting to collect and process transport data.
With the help of such sensor technologies, sensor data analytics can help public transport in people counting, transport planning and/or improvements and real-time operation management. It’s no wonder public transport is the top use case for IoT sensor data analytics.
Of course, due to my bounded rationality, I may have missed more sensor tech and transport use cases (like maybe temperature and gas levels). So there could be more to the story than what has been written here. But it’s a rough idea that should make sense.