XMC2Go Design Challenge: Monitoring of pesticides on smartphones

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
Not applicable
XMC2GO Design Challenge 2014 Italy – Results Report


The project aims to provide an electronic device able to detect the presence of pollutants resulting from the use of pesticides in agricultural areas (vineyards, orchards, harvest field, etc..). The device will serve as a real accessory for smartphones, connected via micro-USB. Once connected to a smartphone, an Android app implemented on purpose will allow the user to check in real-time mode all the data collected by the sensor, and store them on the smartphone's internal memory. It may also be used in urban areas adjacent to the above mentioned agricultural areas in order to provide a global indicator of an excessive or improper use of those materials in the municipality.
The small size of the Infineon XMC2GO board facilitates the portability of this "accessory" for smartphones and, at the same time, makes it user-friendly for all users.
1126.attach

Pesticides are used in various fields of agriculture. There are different types of pesticides and different quantities to be used depending on the type of agricultural product. Some of these pesticides release residues that are deposited on the agricultural product or inside it; some others release a residue that is deposited in the soil, others still release a residue in the air. The aim of this application is to detect just those elements that pesticides release into the air, using specific gas sensors.
By analyzing the most common low-cost gas sensors available on the market, we have identified the set of substances that can be detected:

1127.attach

In order to provide a device capable of detecting those polluting elements is necessary to build an array of six sensors:

  • TGS 822 (methane, carbon monoxide, isobutane, hexane, benzene, ethanol);
  • TGS 2602 (hydrogen, ammonia, ethanol, toluene);
  • MQ2 (hydrogen, LPG, methane, carbon monoxide, alchool, propane, smoke);
  • MQ131 (nitric oxide, chlorine, ozone);
  • MQ136 (hydrogen sulphide);
  • MQ138 VOC (mellow, benzene, aldehyde, ketone, ester).

We reserve a more accurate study on pesticide residue to be released into the air, verifying the presence of other pollutants. Pesticides are used in different cultures and can release thousands of substances in the air. Our aim is to find a finite set of elements that lead us back to most pesticides.


Hardware
The hardware design shall take into account the features provided by the board, and provide compatibility with the power supply and interfacing schemes. For this reason we decided to adopt the Infineon XMC2GO board, which well fits the project’s hardware needs. Let's summarize the main features provided by the XMC2GO board:

  • small size: 14 x 38.5 mm;
  • CPU: XMC1100 microcontroller based on ARM-Cortex architecture (size: 4 x 4 mm);
  • Flash memory: 64 Kb;
  • RAM memory: 16 KB;
  • frequency: 32 MHz CPU clock, 64 MHz timer clock;
  • power supply: through USB Debug Probe (J-Link) or through 3.3 V external power supply;
  • connectors: two 8-pin headers.

1128.attach

The XMC2GO features a very small size (just 14 x 38.5 mm), a key factor for this project since it allows achieving a compact system with a small footprint, easily interfaceable with a smartphone. The power supply has been designed for working with 3.3 V voltage level in order to facilitate its connection with a smartphone through the micro-USB interface.
1129.attach

Hardware design
By analyzing board characteristics we can go on with an appropriate hardware design whose main feature is the ability to properly detect the changes in the conductivity of the exploited gas sensors.
A common feature of the majority of the gas sensors is the low selectivity, which means the simultaneous sensitivity to different gases. In order to solve the issue and use these devices in real applications, it is necessary to combine the outputs of many different sensors to obtain, by "pattern recognition" techniques, the selectivity level that a single sensor is not able to provide. It is therefore evident that, in the "electronic noses" field (array of gas sensors), both electronic interface and processing module are extremely important. Moreover, the lack of synergy between sensor and interface does not allow to optimize performance in terms of system resolution capability with respect to different types of gases. Often this limitation is due to the fact that gas sensors are semiconductor devices that perform as resistors whose resistance varies depending on the concentration of the gas, but also as a result of the effects of aging.
The idea behind the front-end sensor reading circuit is based on a Wheatstone bridge which allows to obtain accurate values of the variable resistance, represented by the sensor itself. As mentioned before, in order to achieve a more accurate measure it is possible to adopt an array of gas sensors, in this way we can overcome the cross-sensitivity issue. To ensure greater portability, we suggest to use distributed parameter resistors (SMD resistors).
Furthermore, it is quiet hard to find out sensors that can be powered with voltages inferior to 5V. For this reason, for both the input and output analog readout circuit, we can use a level converter in order to connect a device powered at 3.3V with a 5V powered system. An example is provided by the TXB0108 level converter, which is also bidirectional.
1130.attach
Data processing
As regards the processing of data coming from the device, they will be sent in input to a classifier circuit that, thanks to algorithms properly designed to detect harmful substances, will be able to output exactly the level values of the levels of emitted substances. In order to ensure a certain reliability of the output data, it will be necessary to achieve upstream a training system to recognize individual substances through tools such as artificial neural networks. This will lead to classifiers correctly configured to be implemented at the firmware level.

Firmware and IDE
The firmware read the resistance values of the sensors array and process the data according to the above mentioned Data Processing criteria. The IDE used for firmware development is DAVE, developed by Infineon and designed to be used on the XMC family of microcontrollers.

The App
The App will have a dual function: it will work as a simple viewer of the real-time data collected by the hardware, and will allow the storage of such data into the smartphone flash memory together with, possibly, the GPS position and the current date. In this way, the device may be used for the implementation of a real low cost monitoring network, also facilitating the post-processing of the collected data.
0 Likes
0 Replies