Introduction
With the introduction of autonomous vehicles and a demand for "co-bots" that work together with a human to increase both efficiency and accountability, IoT platforms that can be easily modified to perform a wide range of tasks as well as take command from a human are increasingly needed. These robots, whether working with factory workers, leading search-and-rescue missions, or even save lives abroad, need to be connected 24/7 and also allow humans the power to override commands as necessary in an emerging workplace. Thus, we have created what we call the "Drone Carrier", a complete platform that will help workers deliver packages, help police secure areas, and also be a stepping stone to autonomous swarm applications that could help mitigate traffic (when applied to autnomous vehicles), and eventually will be crucial in aiding us in becoming a multi-plantary species. In the scope of our project, we have created a rough prototype to debug and test our eventual development of the platform. Keep in mind this is only V1 (90% completed, with V2 currently in development).
Cellular IoT
The cellular IoT portion is the most important aspect of our drone. Many people have developed drones already, with the Predator Drone from General Atomics being the most famous, and arguably, most high-tech one out there. Consumer drones like the Parrot or DJI Mavic are great for consumer use, but rely on wifi or other forms of communication that may not be always suitable. Thus, the cellular IoT platform comes in. The cellular IoT platform allows the drone to be positioned in places that wifi may not be able to reach, and would be too expensive for radio triangulation. Cellular IoT is the niche that the market has failed to realize may help solve problems that could decrease the cost of industrial drones, while being more robust than drones that run on wireless connectivity. In our project, we have demonstrated using Soracom's Cellular Sim Card by sending commands through Soracom's Virtual Private Gate (VPG). With this, we have demonstrated the ability to remotely control a drone without the need for radio or wireless, allowing our drone to maintain a smaller manufacturing cost than one for radio, but allowing it to be more robust than one with wireless.
Application
Applications of this technology are endless. Not only can our drone be adapted for different mission scenarios, but it can also be expanded upon or detracted, depending on your mission needs. Our drone carrier is currently designed to carry roughly 4 smaller drones, allowing one drone to become five drones which mean search and rescue operations can proceed faster at a multiplicative and exponential rate. Because our model is decentralized, the small drones require low power and report to the mother drone, of which it can be controlled by a person. This is perfect for applications where human control is necessary, but many workers are needed, thus taking advantage of swarm emergent behaviors and Artificial Intelligence to augment the controller's potential. Furthermore, because the bottom of the drone is equiped with two STANAG rails (to NATO specifications), you can easily dequip the drone module and equip another module, i.e. an amazon package or organ delivery system.
Production
For production, we have currently 3d printed the design. As this is V1, we do not plan on scaling up but rather perfect this design so that it may be consumer ready sometime. Otherwise, due to exchangable parts and a standard that we conform to, all specifications for the creation of the drone and replication of it will be easily and parts cheap, allowing for rapid prototyping AND rapid manufacturing.
Conclusion
With the addition of Soracom's Cellular Sim Card coupled with rigerous filters that help stabilize data and incorporate machine learning, all on an adaptable platform, we have considerably revolutionized attempts at creating swarm technologies (rivaling those at DARPA's OFFset program) due to both the low-cost of development (roughly $400 in materials for both mother drones and 4 small drones) and creating an adaptable platform that can quickly switch to match any needs in the workplace. If we do receive a prize, we hope to incorporate it into our existing infastructure to create a better user experience as well as create a more development project. Our next step is to complete development of our V1 drones, and also, if we do win, use the VR headset to create a User Interface in Virtual Reality that will allow a pilot to remotely control a drone if needed, as well as to create a more comprehensive mission control system than one on a website, allowing both 360˚ camera range as well as a more realistic environment. The importance of this product in aiding the industry must be stressed. Not only is our drone platform adaptable, but our software is to. With VR, we can transfer the controls to any robot, in space or on the ground. With our filters, we can effectively run the same code on both quadcopters or fixed-wing aircraft (of which some might need to travel at Mach 1 or more speeds). With our AI deep-learning programs, we can implant our software onto other robots, large or small, without the need to complete redesign a training model from scratch. Our design has been built from the ground up, to be as modular as possible, only requiring small fixes and patches in order to fit new mission restraints. I hope you will decide to choose our project, and if so, we will gladly continue to update you on how your prize has impacted our project and where we hope to take it in the future. Thank you!
Comments