Dairy Analytics

The Big Data and Business Analytics are fore frontier of technologies that are generating huge competitive advantage to the companies. Big Data Analytics are playing an important role in developing new innovative products that are enabling companies to engage their customers in a new and disruptive manner.

According to McKinsey Global Institute Report "Big data: The next frontier for innovation, competition, and productivity" - Big Data Capturing its value: $300 billion potential annual value to US health care and $600 billion potential annual consumer surplus from using personal location data globally.

We specialize in developing Big Data Business Analytics products. Our first Big Data Product, Murai Data Analytics helped save healthcare device makers IT deployment costs in the range of 25% to 40% annually. Next, we applied Big Data technologies to Connected Dairy. Connected Dairy Analytics applies the Big Data technologies, Internet of Things, and Smart Dairy to help improve Dairy operational efficiencies and improve profitability. Connected Dairy, importantly, helps the Dairy operations in proactively identifying Heat Stress and Bovine Respiratory Disease.

Connected Dairy Analytics will harvest huge operational efficiencies, cost savings, and actionable insights to address dairy cattle related critical issues. Connected dairy, importantly, is a data enabled insightful tool that facilitates better management of Dairy activities.

Winner of 2018 IEEE Big Data Cup - FEMH Voice Challenge

Developed Neural Networks that has predicted with 90% accuracy pathological issues in voice detection. The Voice samples were obtained from a voice clinic in a tertiary teaching hospital (Far Eastern Memorial Hospital, FEMH), which included 50 normal voice samples and 150 samples of common voice disorders, including vocal nodules, polyps, and cysts (collectively referred to Phonotrauma); glottis neoplasm; unilateral vocal paralysis. Voice samples of a 3-second sustained vowel sound /a:/ were recorded at a comfortable level of loudness, with a microphone-to-mouth distance of approximately 15–20 cm, using a high-quality microphone (Model: SM58, SHURE, IL), with a digital amplifier (Model: X2u, SHURE) under a background noise level between 40 and 45 dBA. The sampling rate was 44,100 Hz with a 16-bit resolution, and data were saved in an uncompressed .wav format.

Precision Dairy

Key Features of Our Dairy Sensors that make Precision Dairy a reality:

Milk Fever

Milk fever is a disorder mainly of dairy cows close to calving. It is a metabolic disease caused by a low blood calcium level (hypocalcaemia). Between 3% and 10% of cows in dairying districts are affected each year, with much higher percentages occurring on some properties.
Our Dairy Sensors identify Milk Fever (proactive) Symptoms:

Patented Algorithms based on accelerometer & cattle moment activity
Identification of Increased dullness or less activity

Ketosis

Ketosis is a metabolic disorder that occurs in cattle when energy demands (e.g. high milk production) exceed energy intake and result in a negative energy.
Our Sensors identify onset of Ketosis Symptoms:

Patented Edge Analytics that monitors onset of increased activity.
Identification of Yield changes
Production
Feeding

Rumination

Rumination is a window to cow health.Rumination is impacted by external factors as well as internal rumen functioning. Rumination is the regurgitation of fibrous matter from the rumen to the mouth, and the return back to the rumen. This biological process is impacted by the forage’s nutritional and fiber content. Dietary fiber form and digestibility do play a role. Increasing feeding time and dry matter intake has a positive effect on rumination. Yet on any given diet, a cow will control rumination in response to external stressors.Rumination is not only a requirement for healthy cows; it can be a very early indicator of stress or illness. By monitoring decreases in rumination times, management changes can be made to alleviate stressful situations, or to begin treating a sick cow. Stress can be assessed via rumination decreases up to 48 hours prior to any other indicators being evident. Deviations from the herd’s or the individual cow’s baseline rumination time is an early sign of some stressor. But rumination time is not measured by the percentage of cows lying down, Grant cautioned. While dividing the number of cows ruminating by the number of cows in the stall will give you an idea of cow comfort, it does not predict resting time and rumination, or assess changes in time spent ruminating from the baseline.
Our Sensors enable farmers to specify & capture rumination threshold counts:

Lameness

Lameness causes stress, which debilitates and reduces productivity. The financial impact of lameness includes losses from decreased production, cost of treatment, prolonged calving interval, and possibly nursing labor. Loss of milk of 1.7–3 L/day for up to 1 mo before and 1 mo after treatment (because of pain) plus milk discarded because of antibiotic therapy must also be considered. Our Sensors identify Lameness (proactive) Symptoms:

Built-in Machine Learning Algorithms that detect Maleness symptoms
Cloud enabled in-memory grids that compare behavior vs. health ones

Healthy vs. Sick Cattle Detection

Our Sensors identify healthy-vs.-sick cattle:

Patented Algorithms based on accelerometer & cattle moment activity
Behavior Change

Dairy Analytics Platform

Analytics platform consists of two core engines: Machine Learning and Recommendation Engine

The Machine Learning and Rules processing

Rules Processing is performed both at the Edge level and in the Dairy SaaS Cloud. The Edge compute is generally a sensor device and/or a Dairy Mobile App (iOS and Android). The Sensor device module applies streaming rules as the data collected at Dairy. The Edge Rules will alert the Dairy personnel and the rules engine in the Cloud will help provide historical analysis.

Pattern Detection: As part of Dairy Analytics platform, we have developed several Supervised and un-supervised machine learning algorithms. We have analyzed several dairy real-time streams and developed industry standard decision trees that can apply most appropriate proprietary machine learning algorithms ( USPTO - Patent Pending ). Using Machine Learning and Pattern detection, we can correlate the relation between several Dairy optimization factors. For instance, the impact of medication and vitamin intake on the Dairy cattle Milk output. The impact of Seasonal influencing factors on the Health and Milk output.

Historical Analysis: Historical analysis includes the influence of season factors such as Heat Stress(HS) and pneumonia on Dairy cattle output. Moreover, our architecture ties the weather forecast, i.e. weather patterns in coming weeks, to the historical observed behavior on Dairy cattle. For instance, if a Dairy cattle shows illness related symptoms due to sudden temperature changes, our architecture provides actionable insights to Dairy personnel by looking into any sudden Weather changes in near future and immediately alerting Dairy personnel if there any such events.

Recommendation System

Our recommendation system helps Dairy management to handle: Food optimization,Medication Suggestions, Optimal Dairy settings per cattle, and Disease Management. We have developed Dairy recommendation engine by integrating several data sources that take-into both dairy contextual and location details into consideratiion ( USPTO - Patent Pending ).

We apply both content and collaborative (Dairy Farms with similar cattle groups) recommendation systems. The main idea of collaborative recommendation approaches is to exploit information about the past behavior an existing Dairies data for predicting which medication to milk output, medication to disease analysis, and location to social presence analytics.

Smary Notifications: Another important feature of our recommendation system is to prioritize notifications that Dairy management and operation personnel receive and present the notifications such a way that it reduce alert fatigue. We have seen several Dairies small and large where systems (milk production, weather, temperature, health, and other location)generate several notifications per day and if not handled properly can cause work interruption to dairy operations and thus reduce overall productivity.

Social presence Analytics: Our recommendation system applies location based social presence analytics to provide most relevant information to Dairy management. We have seen by applying Social presence analytics not only drastically improves cattle disease management but also enable dairies actively collaborate to the local community.