16th May 2017
CUE Grand Finale £5k Competition Finalists
Discover our CUE 2017 £5K Competition Finalists
Open Diagnostics is a Cambridge-based social enterprise, developing a low-cost ‘programmable’ paper-based diagnostic platform technology, for the detection of crop and livestock viral infections across low income countries. The technology uses ‘cell-free’ synthetic biology, and has already been validated for detection of Zika and Ebola viruses. Our first diagnostic will be for surveillance of Foot and Mouth Disease (Aphthae epizooticae), a globally-distributed livestock virus which can be devastating for rural farmers in endemic regions. The test will simultaneously improve farmers' incomes, by certifying meat as safe for export; and improve the efficacy of national disease surveillance & control programs, which are critical to the management of disease outbreaks. We are supported by the Centre for Global Equality and Synthetic Biology Strategic Research Initiative.
Cambridge Cancer Genomics
Cambridge Cancer Genomics is transforming the way cancer patients are treated through an integrated pipeline to track disease relapse and response to therapy.
Researchers want dynamic medical and genetic data and people want to own their data and learn more about their genome. We have built the heterogeneous platform to allow individuals with genetic data (e.g. from 23andMe) to create a profile and participate directly in ongoing research projects, controlling when and how their data is shared and used. At the same time, we are working directly with rare disease patient groups and researchers to generate, store, analyze, and share dynamic medical and genetic data used in research and healthcare. We believe that creating a secure, patient-centric marketplace for genetic data will make it easier to conduct genetics research and follow-up studies not only in healthcare, but for a wide variety of organisations.
Halo provides wearable pico-solar lighting to ensure every person has access to safe, healthy and affordable lighting. Our approach offers a source of light to the three billion people that don’t have reliable access to electricity and are forced to use dangerous and toxic kerosene lamps and candles.
Our unique technology has led to a substantially smaller device whilst continuing to offer up to five hours of bright light each night. The device can now fit comfortably on the wrist offering portable and wearable lighting. Our devices can allow people to continue to work, study, and socialise at night. The unique portable nature also provides a source of light for people displaced from their homes by natural disasters and civil wars as well as those travelling alone at night. Our devices can provide light to every person at night and offer them the freedom to pursue their aspirations.
Sirona’s goal is to fight gynaecological cancers and STIs through early detection. Sirona is a low cost, non-invasive, at home, rapid diagnostic test that has a unique value proposition. It strives to drive regular reproductive health screening and empower women across generations.
Adjumas is a synthetic biology platform, harnessing an immunogenic protein from a mollusc. Adjumas will develop a new class of adjuvants, molecules that enhance the immune response, to improve cancer vaccines.
We will implement a sustainable and large scale manufacturing process, that will allow us to enhance and tailor the immune response, depending on the selected antigen.
The digitalised world hosts increasingly complex systems, such as production plants, data centres or trading engines. These systems are controlled by several hundreds or thousands of parameters influencing metrics like energy consumption, reliability, and performance. Designing analytical control models for all important parameters is difficult, even if the relevant signals can easily be identified.
Deep reinforcement learning provides an end-to-end framework for learning to control complex systems without requiring to provide a model of how the system works. Yet, as an emerging technology, it is difficult to use in practice. reinforce.io builds open source components to help build practical deep reinforcement learning solutions. On top of our open source framework, we will offer custom integrations, deployments and support.
Millions of Indian children fail to acquire the skills they need to fulfil their potential because they don’t have access to qualified educators. Slate2Learn’s goal is to leverage market forces and data analytics in order to deploy high quality digital tutors at scale at the bottom of the pyramid in India. We offer existing/aspiring tutors the opportunity to start a tech-enabled micro-business in their home, to cater to up to 100 children in their community. For less than $4 a month, children can learn on Slate2Learn’s “digital tutor” daily. Slate2Learn’s proprietary digital tutoring technology combines deep pedagogical expertise with data-driven learning algorithms so as to offer each child a customised learning experience. We opened our first 3 commercial tuition centres in Delhi in January 2017 and we plan to reach 50,000 children by 2019, via a network of 500 digital tuition centres.
logolab.ai is a Software-as-a-Service online logo generator that leverages Artificial Intelligence to instantaneously create original, semantically rich and unique artwork. logolab addresses both the end-users of a logo as well as creative professionals. In the simplest use case, a customer enters a brand name and optional keywords and logolab will automatically and instantaneously generate logos that match the request. As opposed to with human designers, with logolab customers do not have to pay for logo drafts or time spent on generating ideas – they only have to pay if they find a logo they like and want to use.
Icona Diagnostic Systems
Icona Diagnostic Systems develops artificial intelligence algorithms to improve breast cancer diagnosis. Breast cancer is the most commonly diagnosed cancer among women, motivating countries around to the world to implement large scale screening programmes aimed at identifying cancers at early stages in development. But despite the success of these programmes, for every life saved, three women receive false positive results, which can sometimes lead to unnecessary biopsy and treatment.
To solve this problem, we have developed a deep learning algorithm that analyzes and classifies mammograms in order to provide actionable information to radiologists, including lesion location and biological features crucial for making a diagnosis. The predictions produced by our model have great potential to assist radiologists in making better diagnoses, reduce the harms of unnecessary invasive procedures, and reduce the cost of healthcare services