Grant: Engage Grants Program
University of Waterloo
$25,000.00
Automatic discovery of dynamic peer definitions from driving behaviour
FY 2012-13
Summary
With the advancement in sensor and information technology, it is now possible to continuously collect massive
amounts of data in real time when vehicle drivers are driving on the road. This data includes speed, location,
engine revolutions per minute (rpm), acceleration, vehicle health, emissions, and even temperature. The
availability of such a rich dataset allows auto insurance underwriters to understand the risk of their insured
drivers through more direct measurements of actual driving behaviour. Understanding a relevant group for each
driver is important to assess changes in the individual driver's behaviour against their ""peers"". The challenge is
defining ""peers"" in an environment where only individual behaviour is observed. If this challenge can be
overcome, underwriters can then potentially use these dynamically defined peer-groups that contain similar
driving behaviours as an indicator of similar risk characteristics. The objective of this focused project is to
develop a clustering method that is able to automatically discover peer-groups within large sets of driving
behavioural data for the industrial partner, Intelligent Mechatronic Systems (IMS).
IMS focuses on driver-centric solutions and delivers connected vehicle platforms to make vehicles safer,
smarter, and greener. Traditionally, insurance underwriters assess the risk of their insured drivers through the
study of demographic data. However, demographic data does not answer the question of how their drivers
behave on the road, which is a key factor in determining the cost of a car insurance policy. This is where IMS'
platforms can deliver increased value to the insurance carrier. To address the need for intelligent comparisons
and dynamic reference points, IMS needs to move forward and resolve the challenge of categorizing drivers
into peer-groups according to their behaviour. However, this exercise cannot be accomplished solely by human
domain experts due to the sheer size and high dimensionality of data; therefore, an advanced computer-based
application that can automatically discover and define peer-groups is being proposed.
Program
Engage Grants Program
Recipient
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data
As published
Application id
430796-2012
Name nom
Wong, Andrew
Department département
Systems Design Engineering
Organization id
33
Institution établissement
University of Waterloo
Province en
Ontario
Province fr
Ontario
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2012
Competition year année de concours
2012
Award amount
25000
Program id
EGP
Program name en
Engage Grants Program
Program name fr
Programme de subventions d'engagement partenariat
Group en
Research partnerships programs
Group fr
Programmes de partenariat
Committee code
1553
Committee name en
Ontario Internal Decision Committee
Committee name fr
Comité de décision interne pour l'Ontario
Area of application code
801
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Computer software
Area of application fr
Logiciel
Research subject code
2800
Research subject group en
Artificial intelligence
Research subject group fr
Intelligence artificielle
Research subject en
Artificial intelligence (computer vision, use2603)
Research subject fr
Intelligence artificielle (vision artificielle, utiliser 2603)
Partie
1
Nb partie
1
Application title
Automatic discovery of dynamic peer definitions from driving behaviour
Keyword
NSERC
Application summary
With the advancement in sensor and information technology, it is now possible to continuously collect massive
amounts of data in real time when vehicle drivers are driving on the road. This data includes speed, location,
engine revolutions per minute (rpm), acceleration, vehicle health, emissions, and even temperature. The
availability of such a rich dataset allows auto insurance underwriters to understand the risk of their insured
drivers through more direct measurements of actual driving behaviour. Understanding a relevant group for each
driver is important to assess changes in the individual driver's behaviour against their ""peers"". The challenge is
defining ""peers"" in an environment where only individual behaviour is observed. If this challenge can be
overcome, underwriters can then potentially use these dynamically defined peer-groups that contain similar
driving behaviours as an indicator of similar risk characteristics. The objective of this focused project is to
develop a clustering method that is able to automatically discover peer-groups within large sets of driving
behavioural data for the industrial partner, Intelligent Mechatronic Systems (IMS).
IMS focuses on driver-centric solutions and delivers connected vehicle platforms to make vehicles safer,
smarter, and greener. Traditionally, insurance underwriters assess the risk of their insured drivers through the
study of demographic data. However, demographic data does not answer the question of how their drivers
behave on the road, which is a key factor in determining the cost of a car insurance policy. This is where IMS'
platforms can deliver increased value to the insurance carrier. To address the need for intelligent comparisons
and dynamic reference points, IMS needs to move forward and resolve the challenge of categorizing drivers
into peer-groups according to their behaviour. However, this exercise cannot be accomplished solely by human
domain experts due to the sheer size and high dimensionality of data; therefore, an advanced computer-based
application that can automatically discover and define peer-groups is being proposed.