This overview is extracted from the Project Initiation Document.

Project Origin aims to implement a robot system that uses the Xbox Kinect device as its primary sensing mechanism.

The Xbox Kinect is a cheap and highly effective sensor, integrating a RGB camera, depth perception system, and 4-device microphone array. While a lot of amateur and some academic work has put the Kinect to use, it remains to be seen whether the Kinect is the device to take robotics which aim to interact with our environment to the next level in terms of both sensing complexity and cost effectiveness.

The main outcome of Project Origin will be the level of expertise in the way the Kinect sensor works as well as how it can be used; this will be necessary to push forwards with the use of the device in increasingly complex systems.

The deliverable at the end of the project will consist of a demonstration of the use of the Kinect device in the accomplishment of a task set to a robot system in its environment, with a loose goal to push the boundaries as far as possible in the time available.

The project will be managed through the use of standard software tools combined with a project website, which will act as task management and organisation as well as providing a portal for interested parties in the local and global community to participate in, find out about and interact with the project.

The overall project plan consists of phased development of software and hardware systems in parallel. This will enable the hardware specification to be finalised following the implementation of the software, which will govern the hardware components required. Preceding the development phase is an investigation and research phase, providing the knowledge and academic background for the full requirement specification and implementation work. The project close completes the documentation and deliverables.


The Xbox Kinect device has been heralded for its relative complexity as a sensor, while maintaining a low cost through mass-market production techniques and high global sales for its primary purpose. Additionally, the device has been extremely popular among the “hacker” community, utilising the device for all kinds of extra applications. This project will evaluate the Kinect device’s viability for the purposes of a sensing device for robot systems.

Often software for robot systems are developed as bespoke packages for a particular application, while implementing common practices and procedures. Such effort is often a waste. Through the development and reuse of generic components and architectures where possible and applicable, this project will provide applications which may be used in later systems with low implementation and integration costs.

Academic objectives and outputs

The primary objective is to implement the required software in order to use an Xbox Kinect as a sensor data source for a robot system; then make use of the data it provides for mapping and planning purposes.

Additionally, any software developed to perform this task should be suitably de-coupled and generic, so that additional benefits may be obtained by augmenting or extracting software components at a later stage.

An investigation shall also be undertaken to provide an overview of the existing software available for providing such services, and analysis of which – if any – is appropriate for this task.

Project case

The outcomes of this project set to provide invaluable information to the student and the other interested parties on the suitability of the Xbox Kinect device in cheap yet intelligent robot systems.

The viability of future robot systems depends on increasing intelligence not just in software but also in sensors and smart sensor systems, while also driving down costs. The very nature of the Kinect device makes it a strong candidate to help push this drive forwards, with engineers no longer needing to compromise on sensor data type or accuracy in order to keep costs of a system at a reasonable level. For a robot system to suitably interact with the environment it is presented with, such high fidelity sensors are required; for robot systems to ever become mainstream enough to require such behaviour, their cost must not be prohibitive to the average consumer.

The project also stands to benefit the student in experience of creating a fully integrated robot system, as well as providing a research platform for future work.