The multi-billion dollar residential waste management “industry” (both public and private entities) presents a substantial opportunity for innovation through the application of data-based solutions to collect waste more efficiently and improve utilization of waste collection resources. Significant, yet variable, resource inputs (trucks, labor, and fuel being most directly relevant) offer direct cost savings for municipalities or commercial entities able to gain efficiency in applying those resources.
This segment is ripe for innovation for several reasons. Globally, urbanization continues to be a trend, with increasing populations living in ever closer proximity. The composition of waste continues to change between recyclables, compostable waste, and landfill waste. As such, resources for the collection, sorting, and disposal of solid waste continue to move towards increased categorization. Lastly, the waste management industry appears to be receptive to disruptive efforts, as evidenced by cities such as New York that are undertaking significant waste reduction measures.
Furthermore, solid waste management is a sector that has largely avoided any significant optimization efforts. For example, in 2013 the city of Chicago instituted a simple “grid system” for the deployment of its collection trucks and abandoned its previous “ward-based” system. As a result of this relatively simple change, Chicago was able to deploy 40 fewer trucks per day (320 vice 360) and gained an $18M annual cost savings against a 2013 budget of $166M for the Bureau of Sanitation – an 11% reduction. As impactful as that reduction was, it was the result of an unsophisticated and non-data driven solution which did not (and does not) take advantage of numerous technological tools for further optimization.
Utilize existing data available from GPS and scale sensors on-board collection trucks to collect, analyze, and employ information regarding individual or street level solid waste production to more efficiently employ waste collection resources (trucks, labor, fuel, time, etc). Armed with an informed picture of the specific house, street, or neighborhood-level of solid waste production, which would become more informed over time with ongoing data collection, the public or private solid waste collection entity could then optimize its resource acquisition, retention, maintenance, and utilization. Each entity could optimize for route length, “truck-sized” routes (in pounds of waste), a specific shift length, distance traveled, cost, or other desirable optimums.
Modern collection trucks are currently equipped with on-board scales, GPS systems, and vehicle monitoring systems. The on-board scale is used primarily to help drivers comply with weight restrictions (e.g. small bridges and weight restricted roads) and avoid exceeding the vehicle weight rating. GPS data includes time and location and, in turn, speed and number stops. The on-board vehicle monitoring system provides fuel usage data, engine RPMs, and speed. Data from each of these sensors and systems could be downloaded from each truck periodically for analysis and incorporation. Any trucks lacking these features can be readily upgraded at a low cost. This data, merged in the appropriate way, could produce a data set which readily lends itself to powerful resource optimization algorithms.
Exhibit 1: A theoretical example of one truck’s weight over a work day. Increases in weight allow one to deduce the amount of waste collected at each stop. Merging this data with time-stamped GPS and operational truck data would allow one to deduce the amount of garbage collected at the block, street, or even individual building level (contingent on scale accuracy).
The pilot program would be rolled out in a city with a large concentration of residential neighborhoods that would benefit from increased efficiencies in waste management. The program would begin by ensuring the existing fleet is equipped with the requisite on-board systems to collect data about the quantity of trash picked up at each stop along an existing route, whether at the house, street, or neighborhood-level. During this time, additional data would be collected about the costs of fuel, labor and other operating expenses associated with their route. Over the course of several months, the quantity of trash picked up at individual stops, marked on the GPS system, would be aggregated in a central location. After a sufficient data set is collected, the optimization algorithm would be applied to develop a new route or routes optimized for time, pounds of trash, fuel efficiency, distance traveled, costs, or whichever parameters the municipality or commercial entity seeks to optimize. The benefits associated with the optimized routes could then be compared against the original routes to determine the efficacy of the program.
Currently, the global waste management industry is valued at $240B in 2016 and is expected to grow to roughly $340B by 2024. With this growth will come a dramatic increase in fuel consumption, labor costs and other operating expenses associated with garbage truck fleets. The opportunity to apply data-based solutions for the optimization of routes, fleet utilization, and labor force would potentially provide billions in annual savings for both waste management companies and consumers.
Chicago Department of Streets and Sanitation
The City of New York Department of Sanitation
Solid Waste Management Market Share & Forecast, 2017-2024
Trends in On-board Scale Systems for the Waste Industry
Real-World Activity and Fuel Use of Diesel and CNG Refuse Trucks
Average Fuel Economy of Major Vehicle Categories
The Economics of Electric Garbage Trucks are Awesome