top of page
  • Adwith Malpe

Capstone Project: Amazon Telemedicine Robot

For my capstone project at ASU, I was able to put my technical skills to the test in developing an autonomous vehicle that is capable of transporting medication and resources to different rooms in the ICU of a hospital. The pain point we are trying to address is to reduce the workload nurses and doctors have to undergo when battling Covid-19. As more infected patients flood hospitals, medical personnel are required to work harder and longer hours to assist these patients. The pandemic has taken a physical and mental toll on medical personnel all over the states and continues to threaten hospitals as the number of the infected increase significantly. The purpose of our project is to develop an autonomous system that can be deployed in hospitals that are able to transport resources and medications to patients in different rooms located within the ICU. This will reduce the workload doctors and nurses have as they do not need to worry about transporting supplies on their own. With the deployment of our autonomous vehicles, we will help doctors and nurses conserve their energy and invest their time in tending to patients. This project will promote a more efficient system in hospitals as medical personnel can save their time, energy, and resources knowing that they have a reliable system to depend on.

For this project, I have been assigned to work in a team of five, comprised of Computer Scientists and Computer Engineers. Our project is sponsored by Amazon and they have helped us greatly in supplying the material we need to assemble the robot, while also funding our budget. The overall budget of the project totaled $900, $400 of which we invested in Amazon's AWS DeepRacer and the remaining $500 to access AWS services. AWS services would be used to access RoboMaker and AWS DeepRacer software, which provided me with the ability to develop a simulation environment and deploy a reinforcement learning algorithm used by the vehicle for it to learn how to navigate through a simulated or physical environment.

As this project is currently ongoing, I am learning at an exponential rate within fields that range from leadership to new technical abilities. I am also learning more about scrum management and how to implement the concepts within the development process to improve efficiency and productivity. As the summer semester progresses, each member of the team is appointed as the Scrum Master for each sprint. I was able to work as the Scrum Master for Sprint 1 and break down the tasks on the Product Backlog into smaller tasks with the team that could be completed within our two week sprint. For Sprint 1, I picked up the tasks to learn how to develop reinforcement learning algorithms in Python, how to assemble the hardware of the vehicle, and to learn how to develop a simulation environment in Gazebo. I will talk a little more about these tasks in later posts, but for now, the product backlog items for the sprint were broken down even more and assigned to each team member based on their field of expertise. As the Scrum Master, I led the daily scrum meetings from June 11th to June 21st where I would communicate with the rest of the team about what they have been working on, what challenges have they faced along the way, and what do they intend to complete for the day. I was also responsible for communicating with our product owner, who also served as the Amazon sponsor for our project, about the progress we made so far in developing the product increment, which was a proper simulation environment. I made it a priority to develop a simulation environment for this sprint as the increment so that in future sprints, the virtual car can be tested in this environment before testing in the physical world.

We will be nearing the completion of our first sprint in four days and hope to have our simulated environment increment ready by then. In future posts, I will be explaining the development process and all software that we use for specified intents.

10 views0 comments

Recent Posts

See All


Post: Blog2_Post
bottom of page