Medical-Device-Manufacture-v2

Glassbeam Technology – Innovation at its Peak for Medical Device Manufacturers

Recently I have covered several business topics for our largest market segment – healthcare providers and independent imaging centers – on how Glassbeam can help them improve patient care by improving machine uptime and utilization (by analyzing machine logs and related data such as DICOM, HL7, and so on). With our Glassbeam Clinsights™ application suite, there are significant opportunities for the healthcare industry to improve revenues, reduce costs and realize a clear ROI.

Powered by predictive analytics, we are at the forefront of an innovation that promises to deliver a deep impact in the board rooms of some of the largest healthcare providers in the United States. Behind this stellar technology stack are our robust 100+ machine learning algorithms deployed on remote sites seeking to find the tiniest anomaly in the medical equipment that is causing failure and downtimes, but which stay undetected to the human eye. Review this blog that captures the CT Tube part failure case study to get behind the curtains and know the impact we have made most recently: Real Life Business Impact of AI/ML on GE CT Scanners Using Glassbeam Clinsights

Related Reading:

> Who Owns the Data series: Part 1, Part 2, & Part 3

Evolution of Our Technology: Here’s the Backstory

The DNA of Glassbeam’s predictive analytics platform evolved almost 10 years ago as a solution built for global product companies struggling to get actionable insights from their vast installed base (thousands of machines) generating Terabytes of machine log data per day as ‘Call-home’ or ‘Phone-home’ features in their machines. This value proposition resonates really well with medical device OEMs today such as Siemens, GE, Canon, Hitachi Medical, Fuji.

There are several reasons why we are attractive to the Healthcare market segment, but if I were to highlight one key reason, it would be the foundation of our technology stack that is steeped in innovation, backed by patents leveraging the best-of-breed open source and commercial technologies.

Let me outline the 8 essential parts of our technology stack here

1 — Fastest Time to Market: Our platform’s core capability is to quickly turn vast volumes of machine data into actionable intelligence. This is made possible by applications enabled by parsers that simplify the tasks of specifying, parsing and indexing rules for multi-structured data. The platform includes a machine data DSL language – SPL™ (Semiotic Parsing Language) to stipulate the rules for parsing multi-structured data types. Solutions engineers then use SPL to describe the rules and semantics for parsing. For semi-structured data such as logs, this means taking advantage of the inherent structure in the data without any pre-defined grammars; as a result, no specific understanding of databases or procedural programming is required to get started.

2 — Efficiency in Multi-paradigm Programming: The scalable functional programming language, Scala was chosen in 2013 as the foundation. It is efficient and maintainable compared to Java (1:10) – 10 lines of Scala code can do the same and more compared to 100 lines of Java code. Some of the most scalable businesses such as LinkedIn, Twitter, Netflix, Quora, and so on have chosen Scala as their native code base language.

3 — Building At-scale Playbook: When our architects were given the design criteria in 2013 to build a big data platform that could parse and perform ETL on machine logs pouring in from thousands of connected assets, it was apparent that industry-standard SQL databases will choke, or their price-performance curves will not fit the business model. Therefore, Cassandra was chosen as the NoSQL technology as it is exceptionally scalable and could provide high availability and low latency without compromising performance on writes.

4 — Architecting for High-Velocity Data: As a purpose-built analytics database designed for big data workloads where speed and scalability is critical, Vertica offers us advanced analytics functions and in-database machine learning. Glassbeam has yet to reach the maximum potential to use Vertica’s ML capability, however, our 2020 roadmap will roll out initiatives to complement Vertica ML with the ML Lib from Apache Spark’s architecture that is already in place.

5 — Resilient Computation Engine: The big data distributed processing framework, Apache Spark works really well with Scala, Python (our native codebase) and also supports SQL, streaming data, machine learning, and graph processing. It is a perfect complement to Glassbeam SCALAR that is a distributed computing design that allows us to access processed logs from hundreds of nodes in the cloud, without compromising performance and latency.

6 — Enterprise-Class Search Platform: Our search engine foundation data store is powered by Apache SolrCloud. It gives us a highly flexible distributed data processing engine that facilitates searching and indexing of files on a network. It is a strong fit with SCALAR design.

7 — Real-time Analytics: Kafka was a natural design choice for Glassbeam to decouple log processing from analytics and build scalability with a distributed streaming platform. With Kafka, Glassbeam now allows anyone to write their consumers and listen on the Kafka message bus to extract meaningful information for analytics. Kafka is fast and uses IO efficiently by batching and compressing records as needed.

8 — Rapid Decision Making, Visually: Glassbeam selected Tableau in the early days (2009) when its big data engine was not even born. Tableau’s powerful visualization engine turns our raw processed data into visual components. Tableau on top of Vertica is a great combination of fast query/analytics and insightful descriptive analytics.

These are just some examples of what our stellar engineering team has delivered to date. We continue to innovate and improvise our stack to deliver superior ‘big data’ analytics to large product manufacturers.

The testimony to this innovation is the growth in our customer base in 2019. We have seen exceptional growth in our customer base in the Healthcare market. More and more healthcare providers, imaging centers, and third-party service providers have now adopted our technology stack and have begun using machine data intelligence to redesign their operating models.

The era of Glassbeam Clinsights has arrived.

Stay tuned as we explore the next set of innovations in 2020, esp. as we propel forward in the areas of AI and Machine Learning.

Happy 2020! We wish you a wonderful new year filled with prosperity, joy, and success.

Transform Your Healthcare Operations

See how to gain deeper, clearer insight from your machine data to elevate business intelligence and to improve your operational uptime, utilization, and efficiency.

Contact Us