MIT D-Lab class
Hardware for International Development EC.751 / EC.793 (G)
Team
- Juan Angel Luera
- Layal Barakat
- Olivia McGrath, MIT graduate student
Community partner
Greenfoot Africa, Johnson Jacka, Founder and Managing Director at Greenfoot Africa
Country
Tanzania
Proposed solution
A previous D-Lab team, working with Greenfoot Africa, designed IoT hardware, leveraging an ESP32 microcontroller. For the updated prototype, an STM32 will replace the ESP32, because of the improved robustness and ease of sourcing parts. The benefit of the ESP32 is the built-in wifi module, but this function is not needed. As the Greenfoot team is looking to expand production, a scalable
solution is needed.
The purpose of the integrated IoT hardware is to communicate with the fleet management system so that it can assign drivers to jobs. To do this, the requirements of the IoT system are the following:
- Communicate battery and bike health to the fleet
- management system. This includes current, charge,
- and temperature.
- Provide real-time location data of each bike.
- Remotely switch the bike on or off for security.
- Record weight of the cargo.
- Log information on the bumpiness of each ride.
The goal is to have a board that can be prepared for the upcoming production scale up.
Something else we built was the backend of the website. Since the goal is to eventually be able to use the Zelo bikes akin to an uber you need a centralized data system. We wanted the website to be as simple as possible and quick for deployment, therefore we decided to use Flask, HTMX, SQLAlchemy with Sqlite. These technologies enabled us to create a lightweight app that can easily be deployed. Furthermore, by minimizing the number of languages we reduce the complexity of the site itself. Lastly, using a well known language such as python makes it even more accessible. Our goal with this website is that it can be easily passed down to following teams and even be managed by the people in charge of the start-up.
Problem or opportunity
Greenfoot Africa was founded in 2019 with the goal of transforming the transport of goods in Africa’s urban areas through the increased adoption of clean technology. In 2021, they launched a platform where companies in Arusha, Tanzania and surrounding areas can request delivery or pickup of goods. Over the following 18 months, their team gathered valuable information which led them to develop the ZELO eTrike, a purpose-built electric cargo bike and fleet management system, which is the baseline design for which this work is focused.
The ZELO eTrike, which can carry up to 300 kilos fills a need in the community, where current solutions to transport these loads are to either overload motorbikes and take multiple trips, use a petrol or diesel truck which can carry up to a ton - which is significantly larger than what is needed -, or to leverage public transportation or personal vehicles - the latter of which is restricted to wealthier individuals. It is designed to be manufactured locally, with easy assembly, and it can handle rough roads and diverse climates, traveling over 80 km on a single charge [2].
ZELO eTrike drivers are matched with jobs using an intelligent fleet management and matching system. The eTrike benefits drivers as fuel costs alone of operating a motorcycle or van can be up to 50% of total earnings [2].
Now, drivers are compensated based on distance traveled and payload and do not have to pay for fuel. For the driver-job matching system to function and to determine the appropriate driver compensation, integrated IoT hardware is needed for real-time monitoring and tracking of the bikes.
Next steps
We have placed the order for our prototype PCB, which will be arriving a couple of days after the last class. Once the board and the remainder of our components have arrived, we would like to assemble and test the PCB. Once assembled, we will begin to work on getting readouts from each of the sensors.
The Greenfoot team has stated that they are working to finish two bikes in addition to their one testing platform, so there should be three ZELO bikes available to work on by the time our team arrives. When packing for the trip we would like to plan to bring at least three sets of hardware so that multiple units can be tested at once. Along with the testing hardware, we will determine where to mount the hardware once the bike and the necessary housing to keep components shielded from the environment.
The current PCB design uses a CAN transceiver to interact with the battery. Only recently was the Greenfoot team able to acquire a datasheet for the BMS, so in future PCB iterations, it might be possible to eliminate the transceiver, as well as the temperature sensor. The BMS datasheet shows that it should be able to return a temperature measurement. This would help to further simplify the PCB design.
Finally, additional options for load sensing will continue to be explored when the team receives a clearer picture of how the cargo box is attached to the bike. Load cells with the capability to measure up to 300 kilos, and any load cell with a maximum load of 100 kilos or greater which can be used in parallel, are costly and could require additional parts to be added to the cargo box. Over the next couple weeks, we will explore the potential to create our own load cells, customized for the setup, or the use of thin film strain gages on existing parts of the cargo box.
More information
Hardware for International Development EC.751 / EC.793 (G)
Contact
Heewon Lee, MIT D-Lab Lecturer and Research Associate