RFID Data Cleaning for Shop Floor Applications
ABSTRACT In several case studies we found that shop-floor
applications in manufacturing pose special challenges to cleaning RFID data. The underlying problem in many scenarios is
the uncertainty about the exact location of observed RFID tags. Simple filter
s provided in common middleware solutions do not cope well with these challenges. Therefore we have developed an approach
based on maximum-likelihood estimation
to infer a tag's location within the reader range. This enables improved RFID data cleaning
in a number of application scenarios. We stress the benefits of our approach along exemplary application scenarios that we
found in manufacturing. In simulations
and experiments with real world data we show that our approach outperforms existing solutions. Our approach can extend RFID
middleware or reader firmware, to improve the use of RFID in a range of shop-floor applications.
SourceAvailable from: Fusheng Wang
Conference Paper: Efficiently Filtering RFID Data Streams.[Show abstract] [Hide abstract]
ABSTRACT: RFID holds the promise of real-time identify- ing, locating, tracking and monitoring phys- ical objects without line of sight, and can be used for a wide range of pervasive com- puting applications. To achieve these goals, RFID data has to be collected, filtered, and transformed into semantic application data. RFID data, however, contains false readings and duplicates. Such data cannot be used directly by applications unless they are fil- tered and cleaned. While RFID data often arrives quickly and is in high volume, its de- tection usually demands efficient processing, especially for those real-time monitoring ap- plications. Meanwhile, the order preservation of RFID tag observations are critical for many applications. In this paper, we propose several effective methods to filter RFID data, includ- ing both noise removal and duplicate elimi- nation. Our performance study demonstrates the efficiency of our methods.Proceedings of the First Int'l VLDB Workshop on Clean Databases, CleanDB 2006, September 11, 2006, Seoul, Korea (Co-located with VLDB 2006); 01/2006
[Show abstract] [Hide abstract]
ABSTRACT: Radio Frequency Identification (RFID) is set to revolutionise industrial control as it holds the potential to simplify and make more robust the tracking of parts or part carriers through manufacture, storage, distribution and ultimately the supply chain. RFID control is based on unique RFID transponder tags being attached to parts and used to identify the part as it moves through the factory or warehouse. Although RFID dramatically simplifies the process of tracking parts, there are certain situations that can lead to uncertainty about the true location of the part. This paper looks at two such situations: a robotic storage stack and a medicine cabinet. Both cases of uncertainty are successfully resolved by using a statistical filter. This work may lend itself to extensions and generalisations using Partially Observable Markov Decision Process (POMDP) models.
[Show abstract] [Hide abstract]
ABSTRACT: Radio Frequency Identification (RFID) has recently received a lot of attention as an augmentation technology in the ubiquitous computing domain. In this paper we present various sources of error in passive RFID systems, which can make the reliable operation of RFID augmented applications a challenge. To illustrate these sources of error, we equipped playing cards with RFID tags and measured the performance of the RFID system during the di#erent stages of a typical card game.