Innovation Pilot Theme: Smart Data Innovation
In almost everything we do, we generate data. Data can be incredibly useful and commercially valuable, but there is more to its application than the products that seem to be following you around the internet. The projects within this theme seek to generate, analyse and apply data in a way that is accessible, insightful and actionable. Encompassing areas as diverse as drones and agriculture; wearables and running shoe retail; Artificial Intelligence and Language Processing; and city wayfinding and the sensory experiences of people with Autistic Spectrum Condition, these Innovation Pilots will getting smart with big data.
Automated Aerial Crop Inspection
Aerial imagery is a useful tool for gathering intelligence about crop health. Until recently, satellite or manned aircraft were the only options for capturing this imagery, but drones now provide a cheaper and more accessible alternative. Currently, this drone-captured data must be manually sifted in a time consuming process, prolonging the wait for actionable insights.
This project will address this problem by applying machine-learning techniques to enable real-time identification of crop problems captured by drone. The anticipated output will be a proof-of-concept demonstrator application, capable of receiving live images from the drone as it flies and identifying potential problems in real time.
Who was involved?
Shoes2Run, Northumbria University, University of Sunderland
What did they do?
This innovation pilot enabled Shoes2Run to further develop a wearable technology sock called ‘Mymo’. Developing the product involved product design, Artificial Intelligence (AI) algorithm development, smartphone technology and big data analysis. Academic involvement helped develop and test the AI algorithm, giving the product much more credibility when taken to market. Mymo is a platform which helps runners to choose the correct type of running shoe but also informs them of new shoe releases and helps them to connect with other similar runners by creating a community forum to discuss specific shoe types, brands and recommendations.
Shoes2Run intend to continue the Creative Fuse collaboration by working with both Northumbria and Sunderland Universities for further AI development and big data analysis, and has been further supported through next stage product development by the Centre for Process Innovation (CPI) and RTC North.
Smart Sensory Spaces
People with ASC (Autistic Spectrum Condition) can be affected by many sensory factors that are not currently taken into account in the design, adaptation and management of the urban environment.
The project will generate an exemplar sensory map within the North East, including shopping streets, public parks and commercial environments – raising awareness, sensitivity and empathy about how autistic people experience the city as individuals. It would introduce and explain techniques and methods for modifying the built environment to improve the day-to-day experiences of people with ASC.
This will be achieved by combining static spatial data about urban environment conditions with additional data generated directly by ASC stakeholders – gathered through georeferenced mobile datasets, wearable technology which monitors physiological data (such as fitbits), social media, and qualitative online surveys.
Sunderland Data Labs
This project addresses critical infrastructure issues and difficulties in the city associated with wayfinding, impacting on commercial businesses. Creatively gathering and using data, the project will create new assets to be used by a range of community and business audiences.
The project will be developed in partnership with Sunderland Business Improvement District (BID), fusing community data mapping activities with creative design, visualisation and fabrication ‘data lab’ events in order to explore and innovate fresh ways to collect, visualise, and share the experiences of individuals and communities within and around the Sunderland BID.
Who was involved?
Nerds with Words (SME), Sunderland University, Teesside University
What did they do?
This pilot aimed to leverage the latest developments in Artificial Intelligence (AI) for text data analysis. Word Nerds focus was to innovate with existing technology aiming to improve the detail of current social listening and review analysis software. Applying approaches such as Deep Learning, Topic Modelling and Natural Language Processing, the goal was to improve automated machine understanding of product/service performance using open data sources such as social media streams and websites. The ambitious goal of creating an AI that found links between words in a corpus of text, visualised the clusters of text and presented them in a 3D map, rather than solely summarising the data. This outcome was more ambitious than originally and anticipated and the Word Nerds team successfully managed to create a working demo of this AI.
Academic input and collaboration have been instrumental in rapidly increasing Nerds with Words’ AI capabilities. The team are continuing to work on the Word Nerds project and would like to engage a PhD student to continue the work that the Creative Fuse Project began.