A new state developed open source framework has been release on Github which allows tracking software developers to compare and contrast their work against state-of-the art industry standards.
The open source software framework has been developed and led by the UK’s Defence Science and Technology Laboratory (Dstl). The project spans across the 5-eyes nations as Canada, Australia, New Zealand and the United States have also committed time and investment from their own defence laboratories.
The framework, named Stone Soup, is a software architecture that allows tracking software developers to compare their code side-by-side against other codes using realistic data. Stone Soup enables developers to compare their work against state of the art algorithms so industry and government agencies can evaluate the code against standard data sets.
State estimation and tracking capabilities are vitally important for the defense sector as this type of software is critical in tracking adversarial missiles, drones and vehicles. The use cases for tracking software is expansive as it can also be used to follow fast moving drone swarms or the movement of makeshift vessels carrying refuges across the Mediterranean.
The increasingly cluttered space above the Earth’s atmosphere is also starting to get more attention as states begin to plan ways to mitigate the risk posed to satellites and global communication infrastructure by space debris left over from previous launches or decommissioned satellites. Tracking software is vital for understanding the movement of this type of debris, which can often be tiny in size.
Professor Paul Thomas, Senior Principal Scientist at Dstl commented in the release that: “We are really excited to be making Stone Soup available to other tracking practitioners giving us the potential to be high-impact in multiple communities. It’s a ‘standard’ platform for tracking algorithm development, and for testing and benchmarking, which will be a huge benefit for the academic and Defence community.”
Stone Soup Open Source Tracking Software Project
The core of the whole Stone Soup project is the framework itself which is free and available on Github.
The framework is the necessary software infrastructure required to build appropriate combinations of components, such as algorithms, sensor models or simulations, taken from a repository.
The Open Source Tracking and Estimation Working Group (OSTEWG) note that: “The framework is designed to mirror the mathematics in the components and be extensible to incorporate new algorithms developed by the research community, as well as data sets from different fields of interest, thus providing the ability to evaluate new algorithms on different problems more efficiently.”
The evaluation is undertook by an interchangeable metric module. A simulation module is also available in which algorithms can be evaluated against simulations generated by a simulation module. Yet, algorithms are not considered ‘monolithic code blocks’ by the framework and subcomponents of the algorithm are able to be interchange for other ‘functionally equivalent approaches’.
OSTEWG state that: “This is achieved by constructing a standard interface between subcomponents, using inheritance where necessary, such that the calling routine is agnostic to the approach within the code block. The framework is coded in Python 3.0. Python’s class-based, object-oriented, duck-typed paradigm elegantly enables this approach.”
While the project is being spearheaded by state defense laboratories it does have many civilian or commercial implications as tracking software is a key component in autonomous vehicle development and commercial drone delivery projects.
Professor Paul Thomas concludes that: “Before this, it could have taken months, even years, to learn the detailed mathematics of tracking. This is a fantastic tool with so many benefits; I hope lots of experts can join us in using and contributing to this framework.”
“The framework is in its infancy but the long-term aim is to save lives by having data that can accurately track adversaries, giving commanders in the battlefield full situational awareness. It’s an accelerated learning aid for people who are just coming into this area too.”