Project Everest

Work Update

Automating ERS: Uber for garbage tuk-tuks

Chris Swanepoel
Chris Swanepoel | Jul 11, 2017 | in Knowledge Base

ERS Cambodia is on the hustle! With a successful prototype out in Puok Village and progress motoring on in refining both the front and back ends of the business, our team is embracing moonshot thinking to try optimising the ERS collection service, reduce costs and create a more attractive service.

In conjunction to processing payments with Wing (as detailed previously), we’re looking at incorporating business automation systems that will streamline the back end, give greater remote control and provide greater flexibility for the customer.

A system like this can make the service much more modular, allowing collection bags to be diversified to our five bag model that includes separate collection for plastic, cans, paper and cardboard, glass and general waste. This also reduces labour for sorting from the two bag model.

There are two avenues to explore:

1. Using an automated text system

2. Using an integrated app ecosystem like the Uber model

Included is a flowchart that describes how a texting service would operate, however it is easily adapted to using an app and replacing unique numerical codes with QR codes that can be scanned by even the most basic smartphones, the prevalence of which we will establish in our next round of empathising on Friday.

We explored the idea of RFID tagging products, however the supporting infrastructure is much more limited compared to a simple QR code scanning system.

It is also worth noting we are looking to establish a relationship with a local rice bag factory who may be interested in buying our plastic to make new bags and see potential to customise bags through them using screen printing techniques to brand and code bags.

By having a threshold line on the bag that waste must reach before contacting, we are given a time window to wait for enough jobs to accumulate in the same area before dispatching the driver. This window can be flexible and, using machine learning, a limit to waiting times can be established for each of the waste types potentially at each household, allowing for maximum collection with as few trips as possible.

Is anyone aware of an existing platforms or technologies we can tap into? We’d love to fit this into our business model since reducing costs on at the collection end is becoming a priority for our project.

Tagged users
edited on 11th July 2017, 15:07 by Chris Swanepoel

Patrick White Jul 27, 2017

Hi Chris,

This looks awesome, the WISI team in Timor worked with Timor Telecom, our local Telco, to set up an sms system for plumbers.
We were able to connect this to a computer so this could tie in with your machine learning idea.
I would recommend getting in touch with the engineers at Cambodia Telecom (or equivalent) and see what they can offer!

Reply 0

Darcy Connaghan Jul 2, 2018

Status label added: Work Update

Reply 0

Andrew Vild Jul 3, 2018

Status label added: Proposal not viable

Reply 0

Samantha Fisher 2 months ago

What an impressive way to reduce cost and also improving the service , I am sure QRtiger have suggested you this enhancement as l was also struggling first to improve my system but now my profit margin is high with very less complains.

Reply 0