1 DYNamic Acoustic MAPping DYNAMAP Identification of failure

1 DYNamic Acoustic MAPping DYNAMAP Identification of failure markers in noise measurement low cost devices Luca Nencini (Blue Wave Srl) Alessandro Bisceglie (University of Milan – Bicocca) Patrizia Bellucci (Anas) Laura Peruzzi (Anas) Luca Nencini Internoise Hamburg – August 2016

2 INTRODUCTION Blue Wave Srl is an italian startup company involved in the development of the low cost noise monitoring network in the Dynamap EU project. Staff working in the Dynamap project: Luca Daniele Michele Luca Nencini Internoise Hamburg – August 2016

3 INTRODUCTION The Blue Wave company operates in the field of: Electroacoustics and loudspeakers production Customization of web interfaces for noise management Noise monitoring services/systems Luca Nencini Internoise Hamburg – August 2016

4 INTRODUCTION Noisemote low cost noise pervasive monitoring system www. noisemote. com Luca Nencini Internoise Hamburg – August 2016

5 INTRODUCTION Noisemote low cost noise pervasive monitoring system www. noisemote. com Luca Nencini Internoise Hamburg – August 2016

6 INTRODUCTION Noisemote low cost noise pervasive monitoring system www. noisemote. com Luca Nencini Internoise Hamburg – August 2016

7 INTRODUCTION Some other activities carried out by Blue Wave are related to: Alfa Absorption measurement ISO 13472 Noise source localization and classification Tyre – road noise measurements CPX, SPB - ISO 11819 Luca Nencini Internoise Hamburg – August 2016

8 INTRODUCTION The tasks currently carried out by Blue. Wave in the Dynamap project are: - design and build reliable low cost noise monitoring stations with data transmission - design and setup an online infrastructure for data collectioning Luca Nencini Internoise Hamburg – August 2016

9 PURPOSE OF THIS WORK The main concern of low cost noise monitoring sensors is the metrologic reliability and stability over time It appears to be easy to connect a consumer microphone to a cellular phone or to one of the many open-source hardware platforms (mini computers) available today Everything seems to work quite well Luca Nencini Internoise Hamburg – August 2016

10 PURPOSE OF THIS WORK But this happen just in an indoor environment or at least in your garden : those devices are not intended to work outdoor ! As this instrumentation is installed outdoor it is exposed to sun, corrosive agents, gases and dust, electrostatic discharge, etc Luca Nencini Internoise Hamburg – August 2016

11 PURPOSE OF THIS WORK Temperature raises up to 70 celsius degrees inside a sensor box left under the sun in august. So industrial grade electronic equipment had to be choosen. Luca Nencini Internoise Hamburg – August 2016

12 PURPOSE OF THIS WORK The most critical part of the monitoring system is the microphone. It would be useful to identify an evaluation method able to determine how the acoustic parameters of the microphones are changed under the influence of physical and chemical agents found in urban and suburban environments. Furthermore, a remote diagnosis method of the microphones health status should be defined, so they can be promptly replaced at the end of their life Luca Nencini Internoise Hamburg – August 2016

13 DESCRIPTION OF THE EXPERIMENT A group of 4 identical MEMS microphones and a group of 4 identical electret microphones were tested for 24 hours to check their resistance to extreme temperature and humidity conditions A salt spray chamber was used, opportunely tuned to provide high temperature and humidity Before and after exposing the sensors to the salt spray cycle, acoustic measurements were carried out. Measurements were made inside an insulated acoustic chamber, available at ANAS Road Research Centre Luca Nencini Internoise Hamburg – August 2016

14 DESCRIPTION OF THE EXPERIMENT This procedure involved the accomplishment of hard stress conditions with the aim of inducing the damage of microphones under test Luca Nencini Internoise Hamburg – August 2016

15 EXPERIMENTAL SETUP The ANAS salt spray chamber was used to simulate corrosion in extreme conditions (high temperature at 40 degrees, 95% humidity, sodium chloride 5% concentration) 4 MEMS In order to accelerate any ionic corrosion process, microphones were kept powered 4 Electret Luca Nencini Internoise Hamburg – August 2016

16 EXPERIMENTAL SETUP Floor noise and Sensitivity vs frequency were measured before and after the salt spray exposure. Tests were done at ANAS experimental road center in Cesano. Microphone Signal conditioner Loudspeaker Recorder Luca Nencini Internoise Hamburg – August 2016

17 CHECK OF MEASUREMENT CONDITIONS The measurement equipment (loudpspeaker) and the microphones were assembled and disassembled five times in order to estimate the uncertainty associated to the assembling procedure As it can be seen, the error due to the microphone re-positioning is neglectable for our purpose Luca Nencini Internoise Hamburg – August 2016

18 CHECK OF MEASUREMENT CONDITIONS In order to assess the influence of the external noise on the floor noise evaluation, a residual noise spectrum measurement was made inside the test chamber with class I instrument. An overall value of 17 d. B(A) was found. Such a value is sufficiently low to avoid the influence of external noise events in the floor noise measurements. Luca Nencini Internoise Hamburg – August 2016

19 RESULTS Loudspeaker - microphone response and floor noise ELETRECT MEMS before after 5 d. B division Luca Nencini Internoise Hamburg – August 2016

20 RESULTS Plot of microphones showing the biggest differences before/after ELETRECT SENSITIVITY MEMS SENSITIVITY 5 d. B before ELETRECT FLOOR after MEMS FLOOR 5 d. B Luca Nencini Internoise Hamburg – August 2016

21 RESULTS A particular behaviour was observed in the waveform associated to the floor noise in the MEMS. Very small high frequency spike occurrences have been observed in the signals achieved from the MEMS microphones after the aging process. 60 seconds wave Luca Nencini Internoise Hamburg – August 2016

22 RESULTS The effect is appreciable when analyzing the audio signal, but disappears when signal integration is performed to calculate Leq values. Luca Nencini Internoise Hamburg – August 2016

23 RELIABILITY TESTS Microphone Test of Blue Wave microphones thermal response were executed using ANAS thermal chamber in order to evaluate the influence of temperature on measured noise level Luca Nencini Internoise Hamburg – August 2016

24 RELIABILITY TESTS A 0. 2 d. B(A) deviation is shown in the range 0 – 50 celsius. Luca Nencini Internoise Hamburg – August 2016

25 DYNAMAP HARDWARE Hardware specifications: Approximative size: 8 x 10 x 15 cm Weight: about 1 Kg Average power consumption: about 2. 0 W Data transmission capability: 3 G, GPRS, Wifi, Bluetooth, Ethernet Protection: microphone windshield and rain protection Luca Nencini Internoise Hamburg – August 2016

26 GENERAL SYSTEM SPECIFICATIONS Software specifications: Each sensor will produce every second a classified output in the format: 160824101000 64. 2 0 - First value is the timestamp - Second value is the d. B(A) fast level - Third value is a flag standing for “road” or “non road” event A special kind of database suitable for time series management will be used in place of the traditional databases. Luca Nencini Internoise Hamburg – August 2016

27 GENERAL SYSTEM SPECIFICATIONS Luca Nencini Internoise Hamburg – August 2016

28 SOME PREVIOUS TESTS WITH CLASSIFICATION Some results of artificial intelligence use for noise source identification: Luca Nencini Internoise Hamburg – August 2016

29 SOME PREVIOUS TESTS WITH CLASSIFICATION Some results of artificial intelligence use for noise source identification: Luca Nencini Internoise Hamburg – August 2016

30 CONCLUSIONS - In order to guarantee a true low cost automatic noise mapping system, mantainance costs of the sensor network must be minimized. This can be done by producing monitoring stations having METROLOGICAL CHARACTERISTICS STABLE IN TIME AND INDUSTRIAL GRADE COMPONENTS. Lowering too much the hardware quality to achieve lower cost would be detrimental to the success of the DYNAMAP. - Reliability tests and optimizations were done to simulate the ageing of sensors, in order to ensure a long life for the developed hardware. - A possible diagnostic method for MEMS microphones health has been identified. If confirmed, this method could be useful to verify the health of this kind of monitoring stations. Luca Nencini Internoise Hamburg – August 2016

31 Thanks for your attention Luca Nencini nencini@blue-wave. com Luca Nencini Internoise Hamburg – August 2016