Autonomous Navigation and Control of a K-9




The GPS and Vehicle Dynamics Lab at Auburn University is currently working to develop modeling and control algorithms that can be used to autonomously command a dog to reach desired mission objectives. It is anticipated that the dog with augmented controller and sensing devices will prove to be a viable autonomously controlled system for security purposes. The basic insights that we hope to obtain from this effort are how best to design augmentation devices for autonomous control and operations of a dog, determine best practices for training a dog to perform remote tasks and the range/scope of operations for which the augmented dog can be employed. The results produced from this proposal will contribute to the development of methodologies to provide autonomous dogs for a variety of applications such as surveillance, search and rescue, delivery of medical supplies, etc.

Experimental studies have already been performed on the ability of a human to command the K-9 shown in Figure 1 to specific locations as shown in Figure 2. The results from this current work has revealed that dogs can be trained to obey audible direction-oriented commands relayed over long distances consistently and reliably for relatively lengthy periods of time with little to no human contact or intervention. Are current focus is to study and develop mathematical models for the dog’s path-planning capabilities and control system, as well as the dog’s dynamic characteristics. This will then allow for an autonomous controller to be developed to guide the dog to GPS waypoints similar to methods used for unmanned ground vehicles (UGVs).

Figure 1:


Remotely Commanded K-9 with Radio and GPS-INS Receiver

Figure 2:


Experimental Results of Remotely Commanding a K-9