Developing a Smart Wheelie Bin

Michael Carr

Overshoot Day

The day when the resources we use in a year exceeds what the Earth can regenerate.

• 19th August 2014
• 13th August 2015
• 8th August 2016?

​

Factors

• Fossil fuels we consume
• Increased food consumption
• Waste management

​

Internet of Things

• Potential economic impact of $3.9 trillion to $11.1 trillion a year by 2025
• Improve resource efficiency and reduce waste as well as convenience
• e.g. Smart energy systems could save $1.2 trillion in US

​

Waste Management Benefits

• Some benefit can be seen in local authority systems.
• BigBelly - A solar powered rubbish compacting bin that is suitable for use in public spaces.
• SINTELUR - M2M sensor device that is capable of determining the filling level of waste in the containers.

Current Household Waste Management Issues

• Pressure to reduce collections to save budgets. Bury council estimate £862,000 saving per year by three weekly collections.
• Confusing schedules on which bin is required which week.
• Often forgotten to be moved for collection (students are prolific offenders).
• Some bins may be overfull, others underfull.

​

​

Possible Solution

• An IoT device retrofitted to a household bin.
• Array of sensors and actuators to monitor fill-level and location and provide feedback.
• Algorithms analysing data captured on servers.
• Smart-phone application to interact with the device.

Reminders

• User indicates storage and collection locations.
• If not moved, reminder sent to mobile phone.
• Ability to cope with confusing schedules.
• Automatic updating of schedule by councils.

In-App Education

• User data stored on a server.
• Algorithms able to identify the frequency of an overly full bin.
• In-app education then provided on which items to recycle.
• Or information on how to contact the council to request an additional bin provided or ask for further information.

Safety Monitoring

• Record of when last moved.
• Record of when last filled.
• If user from an at risk category, e.g OAPs, & no movement recorded.
• Contact can be made to check on welfare.

Business Use

• Food related businesses (restaurants, supermarkets etc.) often employ external contractors to collect waste.
• Installing a similar system would allow collections only to be made when required.
• Remote monitoring of the fill level by contractor would allow for more autonomous process.
• Save money and time on collections when not required.
• Improve hygiene and space issues when collections are needed.

The Internet of Bins

• Night before collection, a user with free space is asked:

“Would you like to share this space?”

• If yes, a notification is broadcast to other users of the system within the neighbourhood to make them aware of free space.
• Users receive rewards for sharing their empty space, such as points, badges etc.
• If in conjunction with the local authority, could be tangible, monetary rewards.

Prototype

• Indicate how full the bin is
• Identify whether it is in collection or storage zones
• Remind the user when not in one of these zones at specified time
• Customise settings and view the bin’s status through an Android app

Technical Components

• An Arduino microcontroller in the bin controlling:
• An ultrasonic range finder in the bin that detects the distance of objects from itself
• An Adafruit Ultimate Breakout GPS module (with external antenna) which provides lat lng data
• A Digi Series 2 XBee (XB-Z24) module in the bin that allows communication over Zigbee.
• A Digi Series 2 XBee (XB-Z24) attached to a Raspberry Pi Version 2 for the hub.
• Originally a Samsung SmartThings hub but due to difficulties, swapped to get a working prototype to demonstrate.
• An Android application to request an update and view data.

Order of Communication

INSERT DIAGRAM

Android Application

Reasoning

• Hub used to reduce processing power needed in bin.
• Arduino was chosen due to popularity and ease of development.
• ZigBee was chosen due to high range, low power and mesh network topology.
• SmartThings was originally chosen due to it’s ZigBee compatibility and developer community.
• An Android app was chosen due to existing knowledge of development.
• GPS was chosen as it is able to identify where the bin is at all times, not just in specific locations (like with RFID tags).

​

Power-Saving

• Only XBee module always-on
• Uses 21 mAh, equates to 54 days battery life from 6 x AA batteries.
• When data is received, the XBee module uses a relay circuit to power on the Arduino.
• Once readings are got, or the GPS timeout has been reached, the Arduino powers off.
• This read uses 103 mAh power.
• Battery life will vary based on amount of reads requested.

Proprietary vs Open-Source

• Difficulties throughout due to lack of transparency from Samsung SmartThings.
• Two weeks before final deadline, system stopped communicating with the hub.
• Access to the code that registers devices, or more accessible system logs may have helped solve this.
• To get a working prototype, a Raspberry Pi running a server was used to send a request using another XBee module.
• Ideally, an open-source hub with support for multiple wireless connection standards would be used.

Live Demo