Cloud Connected Delivery Vehicles: Boosting Fuel Economy using Physics-Aware Spatiotemporal Data Analytics and Realtime Powertrain Control

PI: Will Northrop, Mechanical Engineering                                          VPRO image 

Co-PI: Shashi Shekhar, Computer Science & Engineering

Pengyue Wang, PhD student, Mechanical Engineering

Yan Li, PhD student, Computer Science & Engineering

This project is developing technology to improve the fuel efficiency of delivery vehicles through real-time powertrain optimization using two-way vehicle-to-cloud connectivity. The project is led by the University of Minnesota and represents a collaboration between engine research and computer and data science.

Funding is provided by the U.S. Department of Energy's Advanced Research Projects Agency–Energy (ARPA–E) NEXTCAR program.

AVL software used in project. 

Background

Large delivery vehicle fleet operators such as UPS use analytics to assign routes to minimize fuel consumption. Algorithms mine historical data collected from vehicles to determine routes before a driver leaves a distribution center.

UPS has also invested in E-GEN series electric-powertrain vehicles produced by Workhorse Group Inc. that allow pure electric driving for extended periods of time and use a small range-extending gasoline engine generator to charge the battery, allowing longer routes. Workhorse uses a two-way telemetry system—Metron—on the UPS E-GEN vehicles for diagnostics and fault prediction.

However, neither the current UPS routing algorithms nor the Metron telemetry system interact with the vehicle directly to improve the fuel economy in real time. This project will lead to a greater than 20 percent improvement in the fuel economy of a baseline 2016 E-GEN delivery vehicle by integrating routing, two-way telemetry, and cloud computing.

Project Innovation

Our team will integrate the E-GEN vehicles with real-time powertrain optimization and two-way V2C connectivity. Among the innovations in this process:

  • The vehicle's powertrain controller will be pre-programmed at the beginning of a route to optimize efficiency using historical data and known parameters such as terrain, weather, and traffic.
  • A predictive cloud-based approach—physics-aware spatiotemporal data analytics (PSDA)—will be used to control the hybrid powertrain using efficient computational techniques combined with physics models for computational efficiency.
  • Powertrain calibration will be optimized and downloaded to the vehicle using V2C connectivity in real-time during a delivery route, compensating for exogenous parameter changes or unpredicted driver behavior.
  • Fuel economy will be further improved by co-optimizing routing with developed PSDA algorithms.

Approach

The project will implement three interventions on the baseline E-GEN vehicle:

  • Dynamic battery state-of-charge optimization

A known route with pre-determined stops, analysis of historical hotspots on similar routes, and exogenous parameters will be used to determine when to operate the on-board engine-generator and when to most effectively use regenerative braking.

  • Cloud-based PSDA engine optimization

Engine speed and load will be optimized using exogenous parameters and look-ahead data while maintaining battery SOC within an acceptable range. A low-order physics-based model will be used to predict fuel use by the engine generator system at a given engine operating condition subject to environmental boundary conditions like intake temperature and humidity and historical engine performance.

  • Powertrain control and routing co-optimization

Novel routing algorithms will be used with the PSDA powertrain optimization approach to yield the maximum fuel economy improvement.

Potential impacts

  • Improved fuel economy of individual on-road vehicles by 20%
  • An additional 20% reduction in energy consumption of future connected and automated delivery vehicles
  • Accelerated commercialization. UPS, in particular, already uses E-GEN vehicles. Large delivery fleet operators are also heavily invested in data collection for reducing fuel consumption and actively track their vehicles.
  • Greater energy security. Project innovations could lead to a dramatically more efficient domestic vehicle fleet, lessening U.S. dependence on imported oil.
  • Reduced emissions. Greater efficiency in transportation can help improve urban air quality and decrease the sector's carbon footprint.
  • Improved U.S. economic competitiveness. Project innovations would further solidify the United States' status as a global leader in connected and automated vehicle technology, while a more efficient vehicle fleet would reduce energy cost per mile driven.
More information about our Vehicle Powertrain and Routing Co-Optimization research, please visit http://vpro.umn.edu.

Project Partners

 Workhorse logo

 UPS logo