Innovative Projects Realized

Development of Rapid Inverse Kinematics (IK) Modules for Integrated Robot Simulator

Industrial robotics (i.e. using robots for manufacturing processes) is one of the fastest growing industry sectors today. Traditionally, the process to program industrial robots has been laborious and time-consuming, with each robot having to be taken off-line for setup. Offline programming allows the robot users to program, simulate and generate code for multiple robots at the same time without disrupting the manufacturing process. To gain competitive edge in the global market for such specialized robot simulation programs, the industry partner is collaborating with researchers at University of Waterloo to develop rapid inverse kinematics (IK) code that will improve the ability to solve robotics problems on behalf of the user and make industrial robotic programming easier and faster. Users of offline programming solutions are not robotics experts. They are manufacturing trades people (e.g. Welders). Computationally efficient IK codes to be developed through this project will allow for algorithm development that can iteratively find optimized error free robot paths without user intervention.

Forecasting and modeling future call arrivals at Hydro-Quebec’s call centers

The call center managers at Hydro-Quebec (HQ) need to deliver both low operating costs and high

service quality. Their task is made especially difficult because they need to meet an incoming demand

which is typically both time-varying and uncertain. The current techniques of forecasting future call

volumes at HQ are often unreliable, and there is a need for new and more accurate methods.

In this project, we propose addressing the concerns of the call center managers at HQ by developing

new methods of forecasting future arrivals. Our methods will exploit dependence structures in the

arrival counts, such as those between successive days, between different periods of the same day,

and between alternative call types handled at the call center. We will test our models using real-life

call center data which was provided to us by HQ, and will implement the new, superior models in their pre-existing forecasting technology.

The Formation of Devotional Communities in North India

As part of my doctoral studies at McGill, I will conduct archival research on the formation of bhakti (devotional) communities in North India. I will focus on the Ramanandi Sampraday, which formed around the fourteenth-century Hindu guru Ramananda and is still one of the largest ascetic sects in Hinduism today. There are many stories written by and about his followers, including popular poet-saints like Kabir and Ravidas. I will obtain scans and complete translations of historical texts and literary manuscripts written in Hindi, Braj Bhasa (a language related to modern Hindi), and Sanskrit that have not yet been edited, translated, or published. I will collaborate with Professor Pratyay Nath, a history professor at Ashoka University in Delhi during this project, and I will visit libraries and archives in Delhi, Uttar Pradesh, and Rajasthan. I plan to present the results of my research at academic conferences and use it in my dissertation.

Investigation of the joint effects of smoking and breastfeeding on maternal postpartum weight retention

The harmful effects of smoking on fetal development, child health, and maternal wellbeing are well described, making smoking cessation during pregnancy and the postpartum period a global health priority. Although a large proportion of expecting mothers will spontaneously quit smoking during pregnancy, the majority will restart smoking within 6 months post-pregnancy. Efforts to reduce smoking relapse in the postpartum period have largely been ineffective. Longer periods of breastfeeding, however, have been associated with a delay in postpartum smoking relapse and may increase rates of long-term smoking abstinence. The aims of our research are to determine the joint effects of smoking, which is associated with weight gain, and breastfeeding on maternal postpartum weight retention. Postpartum weight retention increases the risk of obesity, placing new mothers at a higher risk for chronic diseases such as diabetes, cancer and cardiovascular disease. As part of a larger longitudinal health study, research outcomes have the potential to generate evidence for clinical recommendations that will benefit mothers and their children.

Prediction models for pain volatility and engagement patterns of mobile pain app users

Chronic pain costs Canadians more than any other chronic condition and affects 1 in 6 Canadians. Scientific publications of data collected from pain management apps are important to help both consumers and healthcare professionals select the right app. Developed by ManagingLife, Manage My Pain is an app-based pain management platform that allows chronic-pain users to track and analyze their pain experiences. ManagingLife wants to understand and predict pain experiences and engagement patterns of its growing user base of 30,000. We plan to employ scientific methods that will help predict the change in pain ratings, i.e. pain volatility of the app users over time. We also aim to predict the level of user engagement in future based on the current usage behavior and users’ profile information such as gender and age. Through this research, Manage My Pain platform will have analytic capabilities that will enable targeted product development, drive retention and engagement for chronic pain patients.

Evidence-Based Dietary Protein Intake in Canadian Hospitals

Doctors are learning that nutrition plays a very important role in patient outcomes. Research has shown that protein is especially critical to patient health and recovery, especially for the elderly and severely ill. The quality of the protein, and the quantity of protein that patients receive makes a big difference. This project will examine how much Canadian health care centres are concerned about the protein their patients consume. The objectives involve collecting feedback from Canadian hospital staff, and reviewing protein research, including a new protein method that the partner organization has developed. The end goal is to use the work from this project to create compliance and quality improvement research trials in the Canadian health care system.

Effects of Geomechanical Heterogeneity on Wormhole Development During Cold Heavy Oil Production with Sand – Phase 2

Canada possesses vast resources of heavy oil, which is oil that is too thick to flow through porous sandstone reservoirs and into production wells at economic rates when conventional operating practices are used. Since the mid 1980’s, heavy oil operators have demonstrated their ability to increase heavy oil production rates by encouraging the creation of porous and permeable zones (“wormholes”) within their reservoirs by allowing sand grains to detach from the reservoir rock and flow into the well (along with the oil). However, in order to improve the efficiency of these operations, a better understanding of the processes controlling wormhole growth is required. The proposed project involves construction of a novel laboratory testing system, and use this system to run experiments that will help us better understand wormholes in heavy oil reservoirs.

Effects of non-native dandelion on the reproductive success of Asteraceae-specialist bees

There is little known about the extent to which native bees can incorporate non-native plants into their diets, and the costs or benefits associated with their use. Many wild bees are dietary specialists, collecting pollen exclusively from certain plant families, and many specialists prefer pollen from the daisy (Asteraceae) family. In our high-elevation study system in the Colorado Rockies, non-native common dandelion flowers in early June, while most native Asteraceae species flower later. The Forrest lab has observed that bee eggs laid late in the season often don’t survive the winter, suggesting that the earlier availability of dandelion could provide a reproductive advantage to those Asteraceae specialists that use it. This study tests the success of bee larval development on dandelion pollen and explores whether those bees that use this early-flowering plant produce more offspring (have greater fitness) than other Asteraceae specialists. Determining the effects of this non-native species on bee reproductive success can help us to better understand the capacity of these native bees to adapt to changing conditions in their environment.

Inference faiblement supervisee de roles semantiques a partir de documentsspecialises

L’extraction d’information est un processus de structuration de I’information pertinente de documents et elle est

un element cle des applications gerees par KEATEXT. Elle est constituee de trois etapes : extraction des entites

nommees, identification de patrons d’extraction pour detecter leur role et fusion des informations des entites

autour d’evEmements. Ce stage est consacre a la deuxieme etape, celie de I’annotation semantique des

documents avec des roles. Ce probleme est tres difficile a traiter dans Ie cas general, mais nous pensons

pouvoir proposer une solution viable pour des domaines specialises, p.e. medical ou juridique. Les roles

representes par des classes semantiques sont dependants du domaine. Par exemple, pour des documents

composes d’actes notaries de ventes d’immeubles, ces roles pourraient etre Vendeur, Acheteur, Bien_achete,

MontanCde_ Vente.

Neonatal Imitation and Language Preference

This study will examine the relationship between language and communication in newborn infants. In the past, it has been found that newborns are able to imitate the facial expressions of adults. It has also been found that newborns generally prefer the sound of their mother’s native language(s). This study proposes to explore the connection between these two findings by testing whether infants are more likely to imitate an adult who is speaking their mother’s native language. It is expected that infants with monolingual mothers will be more likely to imitate the adult speaking their native language. Newborns with bilingual mothers, however, are expected not to show this bias when exposed to adults speaking both their native languages. The collection of data will take place in a maternity ward of a hospital and will include both monolingual speakers of French and bilingual speakers of French and English.

Effect of Exendin-4 on Cognitive Function in Down Syndrome

Down syndrome (DS) is caused by the presence of three copies of chromosome 21, leading to dementia in later life of the patients. DS patients develop characteristic Alzheimer’s pathologies, including the amyloid-beta plaque formation, tau hyperphosphorylationto form neurofibrillary tangles and abnormal synaptic function and plasticity. Growing evidence has shown that impaired insulin signaling plays a key role in the development of AD and also the pathophysiology of DS. Exendin-4, a glucagon-like peptide-1 receptor agonist has been found to improve glucose homeostasis by promoting insulin signaling. In the current proposal, we attempt to investigate the protective effect of exendin-4 on neurons in DS and the role of insulin signaling in DS-induced cognitive deficits. This study will help us to define the mechanism underlying AD pathogenesis and the prevention and treatment of dementia in DS patients.

Charaterizing Effects of Dye on Microbubble Shells

In photoacoustic (PA) imaging, pulsed light irradiates a sample. Light-absorbing molecules experience a pressure rise and generate an ultrasound wave, from which, an image of the sample is obtained. If the image contrast is insufficient, contrast enhancement is applied by locally introducing stronger light absorbers. In ultrasound imaging (US), shelled gas bubbles are used to enhance contrast. Dyed-bubbles have been proven to be successful PA-US multimodal contrast agents. However, we propose that the process of dying these shells alters the efficacy of the bubbles in US. We aim to study the effects of dye on micro- and nano-sized bubbles on US imaging through computational and experimental methods. Using developed mathematical models, we will estimate the effect of dye on the shell and associated consequences to bubble dynamics. We will validate model results with experimental data. This approach has the potential to optimize bubbles as multimodal contrast agents for PA-US imaging.