From wireless sensors to field mapping: Anatomy of an application for precision agriculture

  • Authors:

  • Alberto Camilli;Carlos E. Cugnasca;Antonio M. Saraiva;André R. Hirakawa;Pedro L. P. Corrêa

  • Affiliations:

  • Agricultural Automation Laboratory, Escola Politécnica da Universidade de São Paulo, Computer and Digital Systems Engineering Department, São Paulo, Caixa Postal 61548, CEP 05424-97 ...;Agricultural Automation Laboratory, Escola Politécnica da Universidade de São Paulo, Computer and Digital Systems Engineering Department, São Paulo, Caixa Postal 61548, CEP 05424-97 ...;Agricultural Automation Laboratory, Escola Politécnica da Universidade de São Paulo, Computer and Digital Systems Engineering Department, São Paulo, Caixa Postal 61548, CEP 05424-97 ...;Agricultural Automation Laboratory, Escola Politécnica da Universidade de São Paulo, Computer and Digital Systems Engineering Department, São Paulo, Caixa Postal 61548, CEP 05424-97 ...;Agricultural Automation Laboratory, Escola Politécnica da Universidade de São Paulo, Computer and Digital Systems Engineering Department, São Paulo, Caixa Postal 61548, CEP 05424-97 ...

  • Venue:
  • Computers and Electronics in Agriculture
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Precision agriculture demands intensive field data acquisition, which is usually done as machines perform field operations. However, more frequent data acquisition and interpretation can be the key to understanding productivity variability. Wireless sensor networks are a new technology that can provide processed real-time field data from sensors physically distributed in the field. This paper describes a simulated application for precision agriculture in which a network of wireless sensors report their measurements to a collector point, where an estimate for the field properties is calculated. Estimation is obtained using the sensor network for processing and transport of the measured data. Centralized and distributed implementations for on-the-go kriging and inverse distance weight procedures are compared considering the influence of noise in the measurements and the in-network coding simplifications. We show that a wireless sensor network application can validate a field estimate constructed only upon local data with less than a 3% loss in precision compared to a centralized approach. We also show how to utilize the communication capacities and processing of a wireless sensor network to create new paradigms for precision agriculture applications, elucidating some of the benefits and drawbacks that arise from this distributed coding approach. Finally, we demonstrate the need to simultaneously engineer the application and the technology knowledge and we show how choices in these two domains can influence the results of the application.