TECH OFFERS

Polyhydroxyalkanoates (PHA) have become very popular as they are bio-based, bio-degradable natural polymers which are UV-stable, able to withstand higher temperatures, showing good resistance to moisture and providing a good barrier for aroma compounds, making them an ideal packaging material. In addition, they can also be used for high value applications such as biodegradable solutions for slow-release fertilizers, pharmaceuticals in general, as well as degradable human implants.Conventionally, PHAs are produced using mainly (food/feedstock) sugars or oils as substrates. In contrast, this offer presents two independent methods using the greenhouse gas CO2 for the production of biopolymers. This allows for the environmentally friendly production of bio-based and biodegradable natural polymers in the sense of a circular economy.

400 million tons of plastics are being produced per year. The majority of it is not recycled, but burned, processed into low-grade plastics or ends up in the environment, where it can take centuries to degrade, accumulate in the food chain and poses an immense concern.To address these issues, standardized and rapid enzyme based tests for the investigation of (bio)polymer biodegradability, and enzymes for the fully natural degradation of different plastics such as PET, PU, etc. were developed. These allow for novel recycling approaches, as well as novel solutions for microplastic in wastewater treatment facilities.

Blockchain-based technology to improve business data management, traceability of information and data security in any sector, establishing an interoperable ecosystem for authenticity and integrity of data. The technology includes a dedicated plug-and-play and cost-effective decentralized application (DApp) built on blockchain for Food supply chains actors in order to address problems such as product counterfeiting, frauds and supply chain fragmentation: For traceability of products and visibility of supply chain from farm to fork To increase transparency and quality control on products and procedures To enable businesses go digital with an innovative, secure and interoperable technology To improve business efficiency through Smart Contract To increase brand reputation, reliability and service quality by establishing a direct channel to final consumers The team also provide developers with a toolkit for software development (SDK) to build new applications (DApps) and an hardware (HW) to software (SW) framework that enables direct integration of devices and machines into blockchain via internet connection (IoT). The team uses their expertise on blockchain and related technologies to build customized solutions for organizations in any industry and integrate digital systems into blockchain through our technology.

There are 2.2 billion blue-collar without a company e-mail address or any other digital communication channel by their employer in many industries such as automotive, pharmaceuticals or logistics. They all want the same: fast, simple, digital communication. They’re kept in the past of poor management communication although all of them have smartphones. Lack of information leads to expensive fluctuation and loss of productivity. The technology is the right tool to bridge the communication gap between physical workers and their employers via a multifunctional HR chatbot. Workers receive real-time, reliable information from their employer and not only verbally from the supervisors. Announcements can be sent out easily to a target group, and the responses are reported live.

This is a tool to translate an English sentence into Malay and vice versa. Developing a translation tool for low-resource languages like Malay has always been a challenge. The main challenge comes from the fact that machine translation systems typically rely on a huge amount of sentence-parallel data, and creating such datasets is an expensive process. By collecting parallel datasets from various sources, the corpus becomes generic and covers both texts and conversations. The second challenge is to train a Machine Learning model. Neural Machine Translation (NMT) is a recently proposed deep learning architecture that has quickly become the standard approach. It offers an end-to-end architecture with better generalization. In the last few years, researchers have proposed many techniques to improve NMT, including work on handling rare words and using attention mechanisms to align input and output words. The translation system utilizes the most up-to-date NMT architecture, namely the transformer net and the seq2seq architecture. OpenNMT-py framework is used to train the model, which is a standard in the MT community for its robust and modular implementation.

Obesity has tripled since 1975 and is projected to hit 15% of the World’s (and Singapore’s) population by 2024. We need to move beyond rewards-based daily steps trackers towards a better understanding of our customers. Public and Personal Health Insurance and Enterprise Industry Health Care costs are rising. Claims are rising, current knowledge of the customer is limited, and customers experience more “stick” than “carrot” when dealing with these industries. Yet insurance providers are looking for a way to better connect to their customers. We are developing an intelligent (machine learning) AI platform that will build on established behavioural science, intervention & change research and back-end data algorithms, to personalise each person’s activity and dietary preferences, as well as their motivational blue-print.

Fine-grained, unobtrusive and individualized monitoring of gym exercises is a high-value problem.  Quantified insights can help users track and improve their own exercise routines and provide corrective postural feedback to prevent injuries. We use a simple, cost-effective sensor, containing only a 3-axis accelerometer and a 3-axis magnetometer, mounted on the weight stack of gym exercise machines, to obtain fine-grained insights into multiple aspects of gym exercise behavior. This is in contrast to prior approaches that either use wearable sensors (obtrusive and requires active user engagement), or video sensing (privacy concerns & inaccurate in multi-user settings).Our approach utilizes the sensor to:•    Identify who is performing the exercise•    Track what exercise the individual is performing, •    Track how much weight a user is lifting•    Identify whether they are doing the exercise correctly.

This Indoor Location System is equipped with the capability of locating any device which is Wi-Fi enabled, in an indoor space where the Wi-Fi infrastructure is already in place. This technology does not demand any additional cost in terms of Wi-Fi infrastructure and also the individual battery power of the client devices. The technology works by analysing the uplink transmissions that any WiFi-enabled mobile device makes to the neighbouring WiFi APs. It provides accuracies typically in the range of 6-8 meters, works with most products from popular WiFi vendors,  and the location information refreshes every 10-20 seconds. Thus, unlike client-side indoor location systems that support only some mobile devices, our technology supports all the Wi-Fi enabled client devices irrespective of their make, model or OS.

Personalised lifestyle analytics has always been of interest to researchers in domains such as healthcare, where researchers may be interested in understanding whether diabetes can be prevented by providing intervention. In the retail domain, researchers are interested in understanding factors that lead to changes in buying patterns. Our technologies dramatically reduce the human effort in which such data capture and analytics require, by relying on automated, smart sensing carried out by personal devices (such as smartphones and smartwatches). In particular, we utilise the inertial sensors and the embedded camera of a smartwatch to capture an individual’s eating behaviour and diet choices unobtrusively. Similarly, we utilise the inertial and Radio Frequency (RF) sensors on a shopper’s smartphone and/or smartwatch to capture their in-store interactions with different products and build deeper profiles for each individual shopper.Current approaches involve either significantly higher manual effort or more extensive infrastructure deployment. For example, for eating analytics, existing approaches require individuals to manually upload pictures of their diet or enter their eating activities into digital journals. For retail, alternative approaches involve the use of in-store cameras and videos, which have privacy concerns and cannot attach an observed shopping profile to a specific customer.