Problem Description
Transportation is a service with a highly increasing demand, given its affordability, comfort, and ease of use. Thus, space becomes an important constraint that has to be taken into consideration. Promoting dynamic interactions between humans and space can help us improve the transportation ecosystem.
The concept of dynamic space is enabled by a vehicle that can autonomously navigate in a geo-fenced environment. Modular pods are developed, which serve different functionalities (for
example, a vending machine or a mobile dispensary for first aid). The autonomous vehicles can dock and undock with these pods, and transport them to the desired locations. They can also
navigate on their own in order to reach and switch between pods.
Our project focuses on the software stack that is necessary in order to achieve this, with emphasis on the behaviors of docking, safety, and navigation. The project aims to develop a set of behaviors for the vehicle which will enable the docking/undocking in a precise manner. Additionally, the robot should exhibit suitable safety behaviors and emergency handling. This set of behaviors will enable the vehicle to transport the aforementioned pods effectively. The resultant autonomy not only saves time and reduces human effort, but also makes space mobile, shareable, and configurable.
Use Case
Shreyans and Sanjana are two graduate students, who are currently working on their capstone project at the University Center. They have their validation demonstration due in the evening. Both of them have been working tirelessly all week, and as a consequence have forgotten to have lunch. They also need a boost of caffeine to keep them awake and to help them focus. However, deviating from project work for even half an hour at this stage could be the difference between a successful and failed project demonstration. Additionally, Shreyans cannot even walk properly due to a heavy workout the previous day.
Realizing that they have neither the time nor the energy to go out and get coffee or food, they turn to the PIX Moving service which transports vending machines on campus. Shreyans opens an app on his phone and requests that a food vending machine is delivered to the UC. The PIX vehicle on campus is alerted of this request and goes to the location of the requested vending machine. Once it reaches the payload handling zone for the vending machine pod, it aligns itself with the pod and follows the path to dock (fig.\ref{fig:use_case}a) with the pod. On reaching the pod it verifies that the alignment is accurate and executes latching with the pod (fig.\ref{fig:use_case}c). It then verifies that the interface between the pod and chassis is active. Having docked with the vending machine, it autonomously navigates to the designated payload handling zone (PHZ) outside the UC and parks the pod there.
Shreyans quickly goes to the pod and grabs the food that both of them want. This saves him the effort of going off-campus in search of food. Now, Sanjana requests a coffee vending machine on the app so that they can have some coffee. The chassis of the PIX vehicle undocks from the food vending machine, plans a path to the nearest coffee vending machine and drives there. A carefree undergraduate student suddenly skateboards in front of the chassis, but the chassis is able to detect the obstruction and stops in time to avoid the student (fig.\ref{fig:use_case}e). Once the passage is clear, the chassis resumes its trip. It reaches the coffee vending machine safely. Following the same docking procedure as with the food vending machine, it docks with the coffee vending machine and drives it to another payload handling zone near the UC. From the array of options offered by the vending machine, Sanjana selects a latte. Upon getting her drink, Sanjana ticks a box on the app which lets the chassis know that she is done using the pod. The chassis then returns the pod to the coffee store. It detects that its battery level is low, and after delivering the coffee vending pod, it navigates to its charging zone.
As a result of the ubiquitous PIX autonomous network, the team is able to have what they need delivered to them effortlessly. The time they save is used productively and they are able to have a successful validation demonstration the next day.