Connected & Autonomous Vehicle: Trip Analysis
Posted by Jai on January 25, 2017
In this post, we will explore to make sense of connected vehicle trip related data and how the vehicle driving information can further be used using data analytics techniques which can be useful for consumer and business also.
We have already have covered what kind of data can be available from connected vehicle in the previous post, IOT Data: Connected & Autonomous Vehicle, Making sense of data. Based on that data we take into account the vehicle driving information available to us.
The information in real scenarios can come from multiple sources. In case of connected and autonomous vehicle, it will come from vehicle itself. In case the vehicle is not connected to such devices, user phone information can be used to identify a trip movement information.
Vehicle Trip Definition
For the connected or autonomous vehicles, the data information we retrieve from the vehicle is the ignition on/off. Now we need to turn this information into meaningful business value trip information.
The functional start/stop information may not be directly mapped to a consumer or business ready definition of a trip. End consumer may switch on/off vehicle multiple times in a span of short period, you can stop your vehicle on a signal, you can also take a break in between your journey etc.
Vehicle trip is logical derivation on vehicle start/stop information. Taking into account the time between consecutive start/stop and small stops during the journey a trip is defined.
Vehicle Trip Analysis
Based on trip information, we will explore what all information can be driven with the data and how it can add business value.
The analysis can be categorised in different buckets,
- Per consumer trip analysis : How each consumer is doing in terms of driving mileage and time for vehicle.
- All consumers travel time analysis: What are the most frequent travel time for consumers
- Total travel behaviour and movement information: How vehicle movement happens
Per Consumer Trip Analysis
In case of connected car solution, from end consumer perspective, lot of analytics information can be provided to the end consumer. Detailed analysis of driving distance/time information and effective travel time can be shared.
Daily Distance Travelled
Detailed analysis of daily distance travelled by a consumer,
As shown above, daily distance travelled by the consumer for a month. Based on the consumer mileage information, consumer can see average distance travelled and on what days vehicle is mostly driven.
Detailed analysis of total vehicle runtime,
As shown above, the total runtime of vehicle daily. How long normally consumer drives on different days, any sudden deviation can alarm the consumer in case driver driven vehicles.
Total idle time for the vehicle,
As shown above, what is normal idle time in each trip for the consumer. Any sharp rise in idle time, can be related to traffic conditions or long idling by the driver.
Daily Travel Efficiency (Runtime vs IdleTime)
Travel efficiency for the vehicle, comparison between total runtime and idle time of the vehicle,
As shown above, how vehicle runtime and idle time give indication of effective driving for the consumer. Based on traffic conditions, which route must be most effective can be analysed further.
Travel efficiency can directly be related to fuel efficiency of the vehicle also. More the idling time, less the fuel efficiency for the vehicle.
Fleet/Consumers Travel Times Analysis
For the entire fleet of vehicles and consumers, added analysis can be useful. For the taxi/cab or shared car solutions this information can be helpful in designing the right business strategy and understand the end consumer behaviour based on day of week, time of day, period of the day and also the city area.
Hour of Day Trips
Indicates which hour of the day the maximum number of trips happen. Which hour of day, the maximum load comes on the system.
As shown above, maximum of the trips happens during early morning hours and evening hours for the business. Using the peak demand of trips and fleet resources, the availability of fleet of vehicles can be managed better.
Day of week trip count
Indicates which day of week the maximum trips happens. Which day of week Sunday/Monday/Friday etc. is the busiest day of the week.
Day of the week, as shown above which day of week on weekdays the maximum number of trips happen in the system.
Further insight into which day of week maximum trips happen, and when the fleet requirements are maximum can be used to further optimise the fleet resources.
Period of day trips count
Indicates which period of the Day morning/afternoon/evening is the busiest time.
As shown, maximum number of trips happen during evening time.
The trip travel time, can be further divided into period of the day, and consumer behaviour can further be analysed based on different periods of the day.
Hour of day vs Trip duration
Indicates how trip duration is co-related with the hour of the day. Which hour of day users like to drive short or long trips.
Indicates which hour of the day the maximum percentage of number of trips of high duration happen. Which hour of day, the maximum long duration trips happen.
From fleet efficiency perspective, short and long duration trips have different significance. As shown above, the peak time of the day can further be categorised based on the hour of the day and trip duration.
Peak hours with long duration trips can be used for consumer targeting.
Distance travelled vs Trip %
Indicates what percentage of trips fall under what distance travelled.
As shown above, the mileage of trip can be directly related to business value. Specific fleet solution, based on short and long duration trips can target either short or long trips accordingly.
Very short trips can be irrelevant for business, but very long trips which are less in number but can be more productive from business perspective. Specific consumer segment targeting can be achieved with the analysis information.
Travel time vs trip %
Indicated what percentage of trips fall under vehicle total travel time,
As shown above, along with the mileage, vehicle runtime information can also be used in similar fashion. Further co-relation between mileage and runtime be used to drive travel efficiency.
Trip average speed vs trip %
Indicates at what average speed vehicle normally travel with. What percentage of vehicle fall under different speed bands,
As shown above, the fleet average speed information can further be used to drive and predict future fleet operations.
Traffic Movement Analysis
Any IOT solution, will have tremendous amount of data processing every second. To maximise the use of such data can lead to lot of additional services to be build for the consumers and market in general.
The vehicle driving data can further be used to help end consumer manage their driving better, and for fleet solution to predict and manage operations better.
Different prediction solutions and predictive analytics can be used using the same.
Check Uber Traffic Movement information on similar model, to display vehicle density in different cities on the world.
Pickups & Drop Offs
Indicates which area has maximum pickup and where users are generally headed to.
As shown above, few specific areas have different density of pickups and city area vise targeting can be achieved from fleet perspective.
Based on the consumer driving habits, the trip destination forecasting can be achieved. We will cover more regarding same in later posts that how we can use existing information to achieve same.
Per City distribution of trips
Indicates which city consumers drive maximum number of times.
As shown above, based on vehicle usage in different city comparative analysis for each city can be achieved. Fleet/Consumer vehicle registrations and where actually the solutions are drives, can give further insights of use cases to business.
Per Weather Trips count
Indicates how weather in general effect the drive habits of people. Which season consumers go out for driving.
As shown above, further insight of user travel behaviour based on different weather conditions can be helpful. In which weather people like to drive more, where users prefer public transport further.
Further analysis, based on different seasons like winter/summer/fall etc. can be added along with weather information.
Have a look at taxi data analysis, Analysing 1.1 Billion NYC Taxi and Uber Trips, with a Vengeance
This post only shows tip of the iceberg in terms of using data to generate some business value. In real scenarios host of services can further be provided to consumer and fleet owners to add more business value based on the data.
Note: All the data set used in the above examples are dummy data, generated to match different analytics features and functionality and nothing to do with real business data or driving patterns.