Bio-Economy and Agri-production: Chapter 15 - Exploiting voice recognition techniques to provide farm and greenhouse monitoring for elderly or disabled farmers, over Wi-Fi and LoRa interfaces

Elsevier, Bio-Economy and Agri-production Concepts and Evidence 2021, Pages 247-263
Dimitrios Loukatos, Athanasios Fragkos, Konstantinos G.Arvanitis

Agriculture is a very demanding sector, in terms of physical condition, and not all people are able to tackle it, in a satisfactory way. The problems become more apparent in case of elder people or people with disabilities. The recent technological advances and the cost reduction in the smartphone industry products and the derived systems can make this problem less intense. This work describes educationally meaningful methods to develop programs that exploit voice recognition engines, based on cloud services, so that a smartphone device can remotely trigger typical farming actions or query the values of several critical parameters of the farm. A step beyond that, the paper also assesses the case when there is no presence of Internet connection to provide access to a remote speech or voice recognition server. In this case, locally running software tackles the voice recognition issues, as embedded boards are now powerful enough to sustain such tasks. In both cases, the node can “speak” reporting the status of a specific element by using speech synthesis methods or by playing prerecorded messages. Not only Wi-Fi radios but also LoRa interfaces are adopted to carry the necessary information at longer distances matching the farm's dimensions. Finally, the paper concludes with fruitful comments on the performance of the presented work and also highlights the priorities for the future.