liu.seSök publikationer i DiVA
Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
A dynamic safety system based on sensor fusion
Linköpings universitet, Institutionen för fysik, kemi och biologi. Linköpings universitet, Tekniska högskolan.
Defence Research Establishment, Linköping.
Örebro University.
2000 (Engelska)Ingår i: Journal of Intelligent Manufacturing, ISSN 0956-5515, E-ISSN 1572-8145, Vol. 11, nr 5, s. 475-483Artikel i tidskrift (Refereegranskat) Published
Abstract [en]

Machines in industry, including industrial robots, have in many cases dramatically reduced the man-made work and improved the work environment. New machines introduce, however, new risk factors. Traditionally machines are safeguarded by means that more or less rigidly separates the machines from the personnel. This works well in many traditional areas, i.e., where industrial robots are involved. There is however a risk that the safety system limits the valuable flexibility of the robot, which can be considered as a quality that tends to become even more valuable in the progress of programming possibilities and sensor technology. This article shows an example how a safety system can be designed to achieve increased flexibility in co-operation between human and production safety strategy. The proposed safety system is totally based on sensor information that monitors the working area, calculate the safety level and improve the system dynamically, e.g., reduce the robot capability in conjunction to the system safety level. The safety system gain information from the sensors and calculates a risk level which controls the robot speed, i.e., the speed is reduced to achieve a sufficiently low risk level. The sensor data is combined with fuzzy-based sensor fusion and fuzzy rules. The safety system is based on sensor information, hence it automatically adjusts to changes in the guarded area as long as the functionality of the sensors is maintained. Finally, we present a system implementation in an industrial robot application.

Ort, förlag, år, upplaga, sidor
2000. Vol. 11, nr 5, s. 475-483
Nyckelord [en]
safety system; sensors; sensor data fusion; fuzzy logic
Nationell ämneskategori
Teknik och teknologier
Identifikatorer
URN: urn:nbn:se:liu:diva-53570DOI: 10.1023/A:1008922330419OAI: oai:DiVA.org:liu-53570DiVA, id: diva2:289377
Tillgänglig från: 2010-01-25 Skapad: 2010-01-25 Senast uppdaterad: 2021-12-16
Ingår i avhandling
1. A Toolbox for Sensor Data Fusion in Industrial Automation
Öppna denna publikation i ny flik eller fönster >>A Toolbox for Sensor Data Fusion in Industrial Automation
1999 (Engelska)Doktorsavhandling, sammanläggning (Övrigt vetenskapligt)
Abstract [en]

The work is focused on measurement for support in industrial automation and especially around industrial robots. The main thread in the work is the combination (fusion) of information from different sensors. The applications are disassembly of electrical motors, sensor fusion in industrial safety applications and a new method for calibration of industrial robots.

When working with worn out products e.g. electrical motors, there are many different sources of uncertainty. To be able to perform operations on the products a method that can work under uncertainty are needed. It can be hard to decide which method works best in a specific situation. To make it easier a sensor fusion toolbox including different methods can be used. Typical data are tested by the toolbox and the method giving the best result is then used in the specific situation. One possible method is fuzzy measures: statistics and operator knowledge are combined forming possibility measures. These are then fused to give a decision regarding which operation to perform.

In the work with disassembly of electrical motors we have studied both commercial sensors, such as vision, accelerometers, current probes and speed counters, and sensors developed at the department, i.e. eddy current probes. One part of the work has been to make the different sensors co-operate, both with each other and with the rest of the system.

Safety is important when working with industrial automation. However, a problem with the safety systems of today is that they reduce the flexibility in the work cell. To avoid that we use new sensing methods, which can monitor the critical area and combine their output by sensor fusion. When an intruder is entering the working area, the speed of the equipment is reduced and when the intruder is close to the equipment it will stop.

To make general off-line programming of industrial robots possible there is a need of high absolute accuracy. Absolute accuracy here means that different robots reach the same position with sufficient precision when the same robot program controls them. The industrial robots of today have a good relative accuracy, in the sense that a specific robot always reaches the same position as before when controlled by the same program. However, the absolute accuracy is not so high due to insufficient robot mechanics. ABB Robotics has, in co-operation with the division of measurement technology, worked with a concept based on a calibration stick equipped within clinometers and a LVDT. The stick measures, from a carefully determined point in the floor and to different robot positions, the distance using a LVDT, and the angle of inclination using inclinometers.

Ort, förlag, år, upplaga, sidor
Linköping: Linköping University, 1999. s. 89
Serie
Linköping Studies in Science and Technology. Dissertations, ISSN 0345-7524 ; 601
Nationell ämneskategori
Annan teknik
Identifikatorer
urn:nbn:se:liu:diva-181866 (URN)9172195568 (ISBN)
Disputation
1999-10-15, Schrödinger (E324), Fysikhuset, Linköpings universitet, Linköping, 13:15
Opponent
Anmärkning

All or some of the partial works included in the dissertation are not registered in DIVA and therefore not linked in this post.

Tillgänglig från: 2021-12-16 Skapad: 2021-12-16 Senast uppdaterad: 2025-02-10Bibliografiskt granskad

Open Access i DiVA

Fulltext saknas i DiVA

Övriga länkar

Förlagets fulltext
Av organisationen
Institutionen för fysik, kemi och biologiTekniska högskolan
I samma tidskrift
Journal of Intelligent Manufacturing
Teknik och teknologier

Sök vidare utanför DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetricpoäng

doi
urn-nbn
Totalt: 198 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf