fleet technology expo - october 2016
TRANSCRIPT
Session Information
Presenter: Lloyd Palum CTO, Vnomics CorpTuesday, October 18
3:10pm – 4:10pmMedium- and Heavy-Duty Session
Equipment Testing: Do You Get Meaningful Results?
This session will explain the importance of validating equipment supplier fuel efficiency company claims, by truly measuring and understanding real world fuel
in/fuel out and not just reading ECM data. The session will be presented by a representative from Vnomics and will focus on the importance of using statistical
process control techniques along with real world fleet performance data to effectively evaluate the impact of equipment on fleet fuel performance.
Equipment Testing: Do You Get Meaningful Results?
Lloyd Palum, [email protected]
http://www.vnomicscorp.com/
What we will cover today1.The importance of accurate measurement2.Isolating the variable effects of the delivery
task and driving behavior from the truck.3.Establishing and tracking truck performance
baselines to use in evaluating any change of operations or equipment and its effect on fuel economy
Good Advice...on testing
“Don’t fall in love with a single test—that holds true for fleets and manufacturers. One test is just that: one test, Multiple tests allow you to look for trends in test data. That’s where we think the most confidence in a [efficiency] number can be found, and then take that number and adjust it to each fleet’s real-world operation.”
Mike Roeth, Executive Director of the NACFELink to Fleet Equipment Article - Is your fleet’s fuel efficiency improving?
Fuel Efficiency Testing
Testing Methods
Computer Model Fluid Dynamics
Wind TunnelTest Track
Road TestingFleet Composite
Real World Factors
WeatherRoad surfaces & grade
Vehicle maintenance levelUtilization profile (duty-cycle)
Manufacturing tolerancesVehicle configuration
Vehicle and component ageDriver behavior
Measurement system precision/accuracy
NACFE Fuel Efficiency ReportLink to NACFE Report
Key Distinction for Today’s Talk
Controlled Tests
• Lock down the conditions (test tracks)
• Explicit assumptions• Set up A/B comparisons• Single point in time• Expensive. Ties up assets that
should be delivering freight• Little correlation to your actual
fleet operations and vehicles
Fleet Composite Tests
• Continuous/Constant• Using real-world vehicles in
real-world operations gives real-world data!
• “What if “ analysis (digital modeling) using real world data.
Benefit of connected trucking [IOT]. GE does this with
locomotives and jets. It's also possible with trucks!
Fuel Measurementprecision and accuracy matter
Fuel Flow vs. ECM Calculations
“fuel consumption data in an ECM is derived from an algorithm and not from actual fuel flow... there is an inherent error with those calculations.” - Yves Provencher Pitt Group http://fleetowner.com/maintenance/certifying-fuel-savings
4 % to 6 % Error
Fleets work hard for percentage point improvements in fuel
economy. Hard to see if you have > 1% error in measurement!
Illustrating the difference...
Actual Fuel Rate
Sometimes higher(more fuel quoted)
Sometimes lower(less fuel quoted)
Eliminate Measurement bias with fuel flow rate
Isolate Fuel Factorsroutes, drivers and trucks
Fuel Economy Factors•The Delivery Task - load, route, traffic,
weather•The Driver - engine control, acceleration,
idling, speed• The Truck - make, model, year, drivetrain
configuration, aerodynamics, tires, maintenance cycle, etc.
The end game is to isolate the truck from these other factors so any tests or changes can
be assessed without bias/influence.
Today’s Example Fleet Dataset
Example Fleet124 Trucks of
various makes models and years. Real
world composition!
• FREIGHTLINERS 2002 - 2017–Cascadia, Columbia
• PETERBILT 2007 - 2016–386, 579
• KENWORTH 2005 - 2016–T880, T680
Fleet MPG Distribution
More than 65,000 Trips
[key on to key off]
Multimodal distribution
MPG - “it depends”
Do you know your trip MPG?
Classification of Trips•IDLE - high idle percentage. A task associated
with yard/dock dwell would fall into this category.
•CITY - frequent stops and less than highway average speed characterize these trips.
•HIGHWAY - infrequent (no) stops and an average speed that is indicative of highway driving.
Isolating the Trip Type
Key part of isolating the delivery task
IDLE1.01 MPG Avg
CITY4.46 MPG Avg
HIGHWAY7.34 MPG Avg
The critical task: Actual vs. Potential
Isolate Driver’s Behavior
• Engine Control• Speed Control• Idling Control
Account for truck configuration
• Engine RPM vs. Differential Ratio at Highway speeds (gearing)
“How do you do that?” Capturing sensor data on every trip and every truck, modeling
(learning) what is possible vs. what is actually happening in each of the above dimensions. This allows us to understand not only the
MPG but also the “Potential” MPG (of course a higher number) if we were able to account for these sources of waste (loss).
How to Isolate the Truck’s Fuel Economy
Model the Waste Factors
“digital twins”
-
SensorMeasurements
Real WorldFactors...
ActualMPG
(measured)
PotentialMPG
(modeled)
Actual (blue) vs. Potential (green)
Green distribution is the fleet’s MPG isolated from driver and
truck configurationeffects on average 0.26 MPG
difference.
Potential MPG - green distribution (waste removed)
Actual MPG -blue distribution
Specific truck and trips
0.27 MPG Diff
0.56 MPG Diff
0.02 MPG Diff
Specific Truck and Trips
Statistical Baseline“digital twin” for each truck
Statistical Control of MPGThe key is to “know” what your trucks are
capable of in real world operations and using that understanding to introduce changes in a controlled fashion to “prove” whether a given change is or is not beneficial for your fleet. No longer relying on anecdotal evidence that may
or may not translate to your trucks or your operations.
Moving Average on 2 TrucksFREIGHTLINER Columbia Eaton Fuller RTLO-16913A Manuals
Introduce change here...
MeasureFuel Before vs. Fuel After
With visibility across the fleet down to a specific truck, at any moment in time you can assess a potential change or react to an unintended consequence on fuel economy by
having a baseline of “true” fuel understanding with which to control your fleet’s fuel efficiency.
Truck B 2017
Truck A 2016
Highway Potential MPG
0.8 MPG Avg.Difference.Ask Why
What we covered todayFuel expense is a never ending consideration in profitably managing a fleet. Having visibility and control over the fuel use on your trucks requires three key factors:
• Measurement must be accurate/precise and fine grained enough to know exactly how much fuel is actually burned/used on each and every trip.
• Isolating the truck from the variable factors of driving behavior and the delivery task allows you to form a statistical baseline of truck performance.
• Creating and maintaining a baseline of statistical control for each truck is the key to being able to assess fuel efficiency over time and under specific changes in equipment or operations.
The outcome: ROI assessment on Fuel Economy investment becomes built into your daily process.
Thanks for listening!