SWTRACK: An intelligent model for cargo tracking based on off-the-shelf mobile devices Rodrigo R. Oliveira a,⇑ , Felipe C. Noguez a , Cristiano A. Costa a , Jorge L. Barbosa a , Mario P. Prado b a Applied Computing Graduate Program (PIPCA), University of Vale do Rio dos Sinos (Unisinos), São Leopoldo, 950, Unisinos Ave., Rio Grande do Sul 93.022-000, Brazil b Taggen RFID Solutions Company, Unicamp Technology Center, Barão Geraldo University City, 100, Bernardo Sayão Street, Campinas, São Paulo 13.083-866, Brazil article info Keywords: Logistic tracking Geofence Location based services Intelligent tracking system GPS Cargo monitoring system abstract Logistics represent the biggest cost for transportation companies. Companies are investing in monitoring and tracking systems aiming at improving services, reducing costs and ensuring the safety in cargos transports. This article presents an intelligent model to obtain the position of vehicles, using off-the-shelf mobile devices named SWTRACK. The proposal allows distributors companies to track loads and to have real-time information about their whereabouts during a travel. SWTRACK presents an automatic travel management to initialize and finish travels without user interaction. Furthermore, it provides a mecha- nism to monitor detours in planned routes and to send alarms notification through mobile devices. To build that mechanism, we studied the Geofence concept and existing techniques to prepare two solutions based on continuous monitoring. Those solutions were compared and discussed. We also conducted a model evaluation using off-the-shelf mobile devices. Moreover, we provide an impact analysis regarding GPS precision and battery power consumption. In the end, we show that the proposed model generates a reliable system for tracking vehicles. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction The biggest cost for transportation companies is the logistics (Huq, Stafford, Bhutta, & Kanungo, 2010). To improve the logistics management, the companies need information in real-time with a high level of precision (Bowersox & Closs, 1996). The information in question is composed of time and location, which can be ac- quired through the Global Position System (GPS) (Hofmann- Wellenhof, Lichtenegger, & Collins, 1997), a technology commonly used in location systems (Hightower & Borriello, 2001). Companies are investing in monitoring and tracking systems aiming at improving services, reducing costs and ensuring the safety in cargos transports (Zhang, Zhao, Yi, & Wang, 2011). Accord- ing to van Dorp (2002), tracking provides the location of where the monitored entity at all times. Shamsuzzoha and Helo (2011) high- light that the use of monitoring and tracking systems is essential to reduce costs, speeding deliveries and even identifying bottle- necks and operational deficiencies. In Brazil, Anefalos (2008) states that the use of those systems has been contributing to decrease the number of cargo thefts. The logistic tracking for delivery networks is an important issue for providing customer service in the trans- portation business (Shamsuzzoha, Addo-Tenkorang, Phuong, & Helo, 2011) and the continuous monitoring and tracking are re- quired for shipments of high value and important cargos (Yang, Xu, & Li, 2010). Currently, mostly the existent tracking systems use techniques of virtual fence known as Geofence (Reclus & Drouard, 2009), which compares the entity position with a predetermined zone or a point of interest, checking if the entity is inside or outside an area. Those techniques do not allow full coverage of the course, making difficult to determine if a truck or another delivery vehicle is traveling in a planned path. Therefore, we need to use an alter- native technique that allows continuous monitoring of travels, obtaining information of probable deviations or even emergency situations. In this article we present the SWTrack, a model that allows companies to track their vehicles and have control over the trav- eled routes. This proposal enables users to identify whether a vehi- cle is following a planned route or deviating from the original path. To accomplish that, we propose the use of off-the-shelf mobile technology to acquire data. Another characteristic of SWTrack is the use of a Geofence technique named Route Adherence (Reclus & Drouard, 2009) to control the travel. Furthermore, we allow real-time travel tracking, through a digital map, showing the loca- tion of vehicles on a web site. This map shows travel information, using colors to identify the travel status. Another feature of the proposed model is that devices can send alarms notification when- ever predefined situations occur. For instance, when a monitored vehicle moves too far away from the planned route. 0957-4174/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.eswa.2012.10.021 ⇑ Corresponding author. Tel.: +55 (51) 3590 8161; fax: +55 (51) 3590 8162. E-mail addresses: rodrigo.remor@gmail.com (R.R. Oliveira), fcnoguez@gmail. com (F.C. Noguez), cac@unisinos.br (C.A. Costa), jbarbosa@unisinos.br (J.L. Barbosa), mario.prado@taggen.com.br (M.P. Prado). Expert Systems with Applications 40 (2013) 2023–2031 Contents lists available at SciVerse ScienceDirect Expert Systems with Applications journal homepage: www.elsevier.com/locate/eswa