|
|
|
# Research
|
|
|
|
**Pre-Deployment Testing of Low-cost Gas sensors.** Nowadays, low-cost air quality sensors are integrated in an increasing number of measurement platforms for air quality monitoring. Calibrating these sensors to reference measurements is however challenging. These sensors typically suffer from cross-sensitivities, poor stability and sensor noise. Information about these limiting effects is often not provided by the manufacturers. Even if the information is given in a datasheet, it is often scarce and reflects sensor performance under laboratory conditions. Neglecting sensor cross-sensitivities and deployment settings usually results in poor sensor performance, frequent calibration necessity and calibration failures. This arises the need for pre-deployment sensor testing under application conditions
|
|
|
|
|
|
|
|
Related publications:
|
|
|
|
* Maag et al. [Pre-Deployment Testing, Augmentation and Calibration of Cross-Sensitive Sensors](./publications#anchor-2016)
|
|
|
|
|
|
|
|
**On-the-fly Calibration** of mobile sensors given a few stationary reference stations (one NABEL and four OstLuft stations in Zurich). Manual calibration of gas sensors is an elaborate and time-consuming task. However, almost all gas measurement instruments require periodic calibration and reference measurements. Therefore, we primarily focus on automatic on-the-fly sensor calibration solutions. We exploit the fact that certain transport vehicles periodically meet each other or pass by static reference stations. Thus, spatially and temporally related measurements are used to adjust calibration parameters, which is necessary to filter out possible sensor aging effects.
|
|
|
|
|
|
|
|
Related publications:
|
|
|
|
* Maag et al. [SCAN: Multi-Hop Calibration for Mobile Sensor Arrays](./publications#anchor-2017)
|
|
|
|
* Hasenfratz et al. [Reducing Multi-Hop Calibration Errors in Mobile Sensor Networks](./publications#anchor-2016)
|
|
|
|
* Saukh et al. [On Rendezvous in Mobile Sensing Networks](./publications#anchor-2013)
|
|
|
|
* Hasenfratz et al. [On-the-fly Calibration of Low-cost Gas Sensors](./publications#anchor-2012)
|
|
|
|
|
|
|
|
**Personal Sensing.** Given the broad availability of personal mobile phones, it is obvious to use these devices to involve the general public into community sensing to sense common air pollutants. We investigate the possibilities and the benefits of such personal sensing.
|
|
|
|
|
|
|
|
Related publications:
|
|
|
|
* Maag et al. [W-Air: Enabling Personal Air Pollution Monitoring on Wearables](./publications#anchor-2018)
|
|
|
|
|
|
|
|
**Route Scheduling** is a problem of selecting a subnetwork of a timetable network to install measurement stations with the goal to optimize coverage of the city given a limited number of measurement stations. Since the measurement stations are equipped with gas sensors which need reference measurements from time to time, we demand that the subset of selected vehicles allows comparing measured sensor values across different sensors. This is possible only if sensing takes place at the same time and same location, meaning that the vehicles selected for the deployment should meet each other from time to time. The set of aggregated meeting points builds a network. Efficient design of this network contributes to the system’s fault tolerance by recognizing sensor malfunctioning, sensor precision loss due to sensor aging, and provides the necessary support for sensor calibration with a reference station. On top of the network of meeting points, we investigate different schemes to periodically calibrate the sensors.
|
|
|
|
|
|
|
|
**Measurement Scheduling.** Given a route plan, a timetable, and a measurement density function. For a given mobile vehicle, regular sampling might lead to a suboptimal coverage of a city. We investigate measurement-scheduling schemes to provide optimal coverage given a route plan and a timetable. Since every measurement degrades the sensor due to involved chemical reactions, the provided problem solution assumes a limited number of measurements per sensor.
|
|
|
|
|
|
|
|
Related publications:
|
|
|
|
* Saukh et al. [Route Selection for Mobile Sensor Nodes on Public Transport Networks](./publications#anchor-2014) |