Christos Goumopoulos

Motivation

Currently, there are few discussions on the integration of biological elements of the real (natural) environment into pervasive computing applications. In this article, we briefly present our research efforts to create digital interfaces to nature, in particular to selected species of plants [1]. Our approach goes beyond the use of sensor networks for environmental monitoring [2] by emphasizing the development of a system architecture that incorporates the plants and associated computation as an integral part of the system, and allows the interaction of plants and artefacts in the form of synergistic and scaleable mixed societies. The ambient intelligence technology is used to encompass plant requirements, by establishing a three-way interaction between plants, people and objects (Figure 1). This approach enables the development of hybrid systems consisting of communicating plants and artefacts in scenarios ranging from domestic plant care to precision agriculture.

Precision agriculture is an agricultural concept relying on the existence of in-field variability across an array of cropping systems [3]. Thanks to developments in the field of wireless sensor networks as well as miniaturization of sensor systems, new trends have emerged in the area of precision agriculture. Wireless networks allow the deployment of sensing systems and actuation mechanisms at a much finer level of granularity, and in a more automated implementation than has been possible before. Sensors and actuators can be used to precisely control for example the concentration of fertilizer in soil based on information gathered from the soil itself, the ambient temperature, and other environmental factors. Incorporating feedback into the system through the use of sensors, actuators, and adaptation algorithms will allow a more fine-grained analysis that could adjust flow rate and duration in a way that is informed by local conditions. One can imagine the use of such precise information in particularly sensitive high value crops such as wine grapes, citrus fruit and strawberries.

The PLANTS system

The main idea behind PLANTS project is to develop a system that produces the optimal growing conditions for a crop, so that crops are kept in the best possible health with the minimum of inputs. It promotes sustainability, because there isn't excessive use of inputs like fertilizer and water. It makes crop management more economic, as well as less damaging to the environment.

The system (Figure 2) uses an infrared camera to scan the entire crop canopy. It can automatically detect when individual or groups of plants are getting too warm. Another sensor detects chlorophyll fluorescence, which tells the system the rate at which the plant is absorbing energy. That reflects the current state of photosynthesis, itself a reflection of the plant’s health.

These sensors communicate their data through specially developed wireless transmitters. The transmitters started out as a Field Programmable Gate Array (FPGA), a sort of universal microchip that can be set to carry out many different functions. Over the life of the project, however, scientists at project partner the Tyndall National Institute in Cork Ireland managed to reduce the essential technology from 100mm FPGAs, to a specialised 25mm module. This system incorporates a wireless transceiver capability with embedded protocol software to minimise power consumption and maximise data throughput. Additional input capability for sensor and actuator integration can be incorporated seamlessly due to the modular nature of the system. This gives it the ability to interface to a broad range of physical and chemical transducers.

The sensors and transmitters are two key elements of the system, but its heart is the management software, called ePlantOS, designed and developed by Computer Technology Institute, in Patras Greece, that gathers, and then acts, on the data operating as a plant/environmental context management system. It can control the deployment of water, nutrients or pesticides, as necessary.

ePlantOS provides an interface with sensors and actuators, maintains a plant-specific ontology and supports distributed resource management with local and/or global decision-making according to criteria. ePlantOS is a middleware system operating according to a peer-to-peer interaction model in which plants being monitored and devices of the system are universally modelled. A modular design allows the replacement of a module without affecting the functionality of the rest provided that the interfaces between them remain consistent. The system incorporates a mechanism for learning rules or parameters to be used in the decision making process, according to historical/logged data.

A set of graphical tools provide the user with a set of useful operations like creating / composing / reconfiguring applications, viewing knowledge represented into the PLANTS ontology, monitoring the plant and environmental parameters and finally managing dynamically the rules taking part in the decision-making process.

Demonstrators

PLANTS developed two demonstrators, one at the beginning of the project in 2003 to prove the concept. The second demonstrator (Figure 3) was presented at the Eden Project, Cornwall, UK end of March 2006 and was the centre point of a special workshop to introduce the technology to experts in the fields of plant science, crop management, microelectronics and software engineering.

The team designed the current system to track high-value, glasshouse crops like strawberries and other soft fruits, but a commercial version could be tractor mounted to survey field crops and, as the costs of production come down, it may be possible to deploy in it in comparatively low-value crops, too.

As well as improving the system, as it exists now, by further miniaturising the equipment, the team also hopes to develop other sensors in a future project. These could look at the levels of volatiles, gases that plants emit when they are under stress.

Conclusions

We are motivated from the fact that plants are truly ubiquitous entities, since they exist in everyday environments. In addition, plants can be used either as "biosensors" or as a "natural" and beautiful interface to services. In the former case, plants could provide additional sensing functionality to be used by humans for example as environmental quality markers; in the latter case, artefacts can use plants as a front end to delivering specific services (i.e. greetings, narration, alarms, etc). Thus, plants, if turned to ePlants, can become part of a range of UbiComp applications ranging from agricultural to domestic.

We have been involved with a facet of precision agriculture that concentrates on plant-driven crop management. By monitoring soil, crop and climate in a field and providing a decision support system that is able to learn, it is possible to deliver treatments, such as irrigation, fertilizer and pesticide application, for specific parts of a field in real time and proactively.