Abstract
Botswana roads' largest assets' estimated value as of 2010 was about 2 billion dollars. Out of the 18,300 km Botswana Public Highway Networks, district gravel road networks are significant in providing access to rural areas where the majority of the population lives. District road managers need intelligent pavement management systems. Although much research has been devoted to performance modeling of pavements, a comprehensive model that could predict gravel loss condition accurately has yet to be developed in Botswana. Accurate prediction of gravel loss condition is important for efficient management of gravel road networks. Moreover, reduction of prediction error of gravel road performance models could conserve significant budget savings through timely interventions and accurate planning. This research developed gravel road performance models using feed forward neural networks (FFNN) modeling technique, which is increasingly used as an alternative to traditional model-based technique to predict gravel loss (GVL) for the first time within a district in Botswana. The input data for the models were generated from the time series triennial condition survey for Botswana carried out in 2002, 2005, and 2008. The expected output, gravel loss (GVL) prediction using FFNN technique gave R2 = 0.94, which outperformed the multiple regression technique of R2 = 0.74 used to compare the models' accuracy. The developed improved intelligent gravel road performance models will give district road maintenance managers information about the gravel loss conditions and equip them with the background needed for sustainable and efficient pavement maintenance interventions to keep the gravel road networks in a good condition in Botswana.
| Original language | English |
|---|---|
| Title of host publication | Airfield and Highway Pavement 2013 |
| Subtitle of host publication | Sustainable and Efficient Pavements - Proceedings of the 2013 Airfield and Highway Pavement Conference |
| Pages | 1358-1369 |
| Number of pages | 12 |
| DOIs | |
| Publication status | Published - Nov 15 2013 |
| Event | 2013 Airfield and Highway Pavement Conference: Sustainable and Efficient Pavements - Los Angeles, CA, United States Duration: Jun 9 2013 → Jun 12 2013 |
Other
| Other | 2013 Airfield and Highway Pavement Conference: Sustainable and Efficient Pavements |
|---|---|
| Country | United States |
| City | Los Angeles, CA |
| Period | 6/9/13 → 6/12/13 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Civil and Structural Engineering
- Building and Construction
Cite this
}
Improved intelligent pavement performance (IIPP) modeling for Botswana district Gravel road networks. / Oladele, A. S.
Airfield and Highway Pavement 2013: Sustainable and Efficient Pavements - Proceedings of the 2013 Airfield and Highway Pavement Conference. 2013. p. 1358-1369.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
TY - GEN
T1 - Improved intelligent pavement performance (IIPP) modeling for Botswana district Gravel road networks
AU - Oladele, A. S.
PY - 2013/11/15
Y1 - 2013/11/15
N2 - Botswana roads' largest assets' estimated value as of 2010 was about 2 billion dollars. Out of the 18,300 km Botswana Public Highway Networks, district gravel road networks are significant in providing access to rural areas where the majority of the population lives. District road managers need intelligent pavement management systems. Although much research has been devoted to performance modeling of pavements, a comprehensive model that could predict gravel loss condition accurately has yet to be developed in Botswana. Accurate prediction of gravel loss condition is important for efficient management of gravel road networks. Moreover, reduction of prediction error of gravel road performance models could conserve significant budget savings through timely interventions and accurate planning. This research developed gravel road performance models using feed forward neural networks (FFNN) modeling technique, which is increasingly used as an alternative to traditional model-based technique to predict gravel loss (GVL) for the first time within a district in Botswana. The input data for the models were generated from the time series triennial condition survey for Botswana carried out in 2002, 2005, and 2008. The expected output, gravel loss (GVL) prediction using FFNN technique gave R2 = 0.94, which outperformed the multiple regression technique of R2 = 0.74 used to compare the models' accuracy. The developed improved intelligent gravel road performance models will give district road maintenance managers information about the gravel loss conditions and equip them with the background needed for sustainable and efficient pavement maintenance interventions to keep the gravel road networks in a good condition in Botswana.
AB - Botswana roads' largest assets' estimated value as of 2010 was about 2 billion dollars. Out of the 18,300 km Botswana Public Highway Networks, district gravel road networks are significant in providing access to rural areas where the majority of the population lives. District road managers need intelligent pavement management systems. Although much research has been devoted to performance modeling of pavements, a comprehensive model that could predict gravel loss condition accurately has yet to be developed in Botswana. Accurate prediction of gravel loss condition is important for efficient management of gravel road networks. Moreover, reduction of prediction error of gravel road performance models could conserve significant budget savings through timely interventions and accurate planning. This research developed gravel road performance models using feed forward neural networks (FFNN) modeling technique, which is increasingly used as an alternative to traditional model-based technique to predict gravel loss (GVL) for the first time within a district in Botswana. The input data for the models were generated from the time series triennial condition survey for Botswana carried out in 2002, 2005, and 2008. The expected output, gravel loss (GVL) prediction using FFNN technique gave R2 = 0.94, which outperformed the multiple regression technique of R2 = 0.74 used to compare the models' accuracy. The developed improved intelligent gravel road performance models will give district road maintenance managers information about the gravel loss conditions and equip them with the background needed for sustainable and efficient pavement maintenance interventions to keep the gravel road networks in a good condition in Botswana.
UR - http://www.scopus.com/inward/record.url?scp=84887324732&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84887324732&partnerID=8YFLogxK
U2 - 10.1061/9780784413005.115
DO - 10.1061/9780784413005.115
M3 - Conference contribution
SN - 9780784413005
SP - 1358
EP - 1369
BT - Airfield and Highway Pavement 2013
ER -