PEDASI – Case Study

The Southampton Research Software Group (SRSG) had only six months to complete the project, and delivered a high-quality, well-documented product that has led to follow-on projects and further funding to enhance its capabilities.

The project to develop the platform was initiated and run by Dame Wendy Hall, Regius Professor of Computer Science at the University of Southampton as part of the wider PETRAS National Centre of Excellent for IoT Systems Cybersecurity programme of work. The aim was to create a data broker service that could collect and store IoT data, and also provide researchers with access to real-world IoT datasets. The platform would be useful in itself and would also be used to explore issues around sharing IoT data, which is becoming especially important in multi-stakeholder environments such as smart cities and smart buildings.

The SRSG was brought in to add expertise on phase two of the project, which had a strict timeline of six months to develop and demonstrate the platform to various key stakeholders in IoT research in the UK. The team had to get up to speed quickly in order to help complete the work before the deadline.

“The PEDASI project was tasked with delivering a sophisticated piece of IoT infrastructure within a very tight deadline. It would not have been possible without the engineering expertise and project management provided by the Southampton RSE Group.”

Prof. Dame Wendy Hall – Principal Investigator, PEDASI

Aastha Madaan, currently working as a Senior Data Scientist at Arup, was employed as a research fellow on the project, and worked with the SRSG as they came on board.

“We were trying to create this platform for sharing IoT data sets. The UK is a leader in IoT, and developing a platform like this was critical to demonstrate the great potential IoT holds to enable cognitive applications across sectors. When we talk about IoT we also have to talk about cybersecurity, and understand how data was collected, and how it can be used,” Madaan says.

The project team was working with industry as well as with researchers, which changed the focus of the development.

We chose to go with the SRSG because we like the way they work. They have an agile way of working, they’ll come back to us with any questions and they were happy to engage with the external partners as well.

Aastha Madaan, Research Fellow – PETRAS

“Cisco and BT were involved, and Google Cloud Platform. Usually, as researchers, the idea is to build prototype systems, just a proof of concept, but here we wanted something that could not only support the researchers at the University of Southampton but also interface with industry,” Madaan says.

With the deadline approaching and work still to be done, the project team considered its options.

“We did look at some start-ups here in Southampton who could have helped us, and they were happy to pick up the project. But we chose to go with the SRSG because we like the way they work. They have an agile way of working, they’ll come back to us with any questions and they were happy to engage with the external partners as well.

“They know how things work when you are constrained by a budget, and their style is just more ‘researchy’ than other providers, as well as being professional. They are researchers at heart, but perfectionists – that’s the best combination you can get!” she says.

The project team began by meeting with the SRSG to discuss what was needed and how to approach the final stage of the project. Dr Steve Crouch, Software Architect at the SRSG, worked closely with Madaan, who was about to leave for another role.

“The idea was for Aastha to hand over her understanding of the project and to help us develop a requirements roadmap, so that I could come up with an architecture for what it would look like, how it would be built and how it would work,” Crouch says.

That handover worked well, making the task simpler for the SRSG.

“It was a really good process. Aastha has a really good understanding of the software process and how we need to go through it, so we spent a good couple of weeks – both asking lots of questions, and increasing my understanding of how the project worked and what it needed to do,” Crouch says.

The handover allowed the SRSG to come up with a software specification that laid out the requirements and identified what could be done in the six months remaining.

The SRSG set up a team and set to work, creating a preliminary version of the software by early November, just three months into the project.

“We had two key goals. The most important was a PETRAS demonstration to show the software was up and running, alongside other PETRAS projects. And the second was a House of Lords presentation where we had to show what it could do,” Crouch says.

The software developed is sustainable, extensible, maintainable and supports demonstrator apps, he says.

“It works with a couple of stakeholder data sets, hosted elsewhere. We showed we can combine data from Cisco, including city data such as cycle route maps with air quality data from Cleanspace, for example,” Crouch says.

“But it’s two pronged – as well as acting as a broker for data, the platform can also be used to explore issues around sharing IoT data. This is a new area, and people aren’t perhaps doing things the way they should. Data is being shared insecurely from lots of devices, and not very clearly labelled as to where it’s coming from. So there are research questions on the provenance of data, on relationships between data, and of course privacy – you need to be sure you’re following data requests where end users don’t want data, or parts of data, shared. So it was a pretty wide ranging project.”

“We had tried to do projects like this before,” says Madaan, “but we were never able to create something like this, a platform that you can expand and create experiments on. All kinds of things are possible now, that previously weren’t possible. The overall approach to the project from the SRSG was something that was really amazing.”