GlideSysLM Platform
Glide GraphWhite Paper
Introduction
Glide SysLM is a cloud-native product development orchestration platform that enhances product development velocity. Most modern hardware products are multi- disciplinary in nature and bring the hardware engineering (electrical, mechanical, optical, etc.) and software/firmware domains together as a system. These products are often connected to a secure cloud adding to product complexity.
Each design domain has separate teams with unique design and IP creation software tools, and they are often disconnected. Product requirements are often managed in Microsoft tools or specialized solutions. Requirements are decomposed based on different design domains and here starts the discontinuity in the product development process. This pain point exists in most companies, be it automotive or medical device or semiconductors. While companies use PLM applications for product development management, these solutions are typically expensive to implement and integration to other applications is an arduous and expensive endeavor.
Our vision is to simplify and democratize platforms and solutions for “systems” oriented product development which has traditionally been the domain of large software companies serving large companies due to complexity and high cost of software and implementation.
Our customers will benefit from rapid time to value and significantly lower cost of ownership due to high configurability and zero customization. Configurability of the platform lends itself to creating industry-specific data models and schema along with business rules which will enable us to expand to several industry verticals.
Concerns
Our first vertical solution is for the semiconductor industry around IP lifecycle management and semiconductor operations bill-of-material.
Today, in several semiconductor companies the system used to manage the design bill of material (IP BOM) is different from the one used to manage the operations bill of material. There is no commercial off-the-shelf platform available that provides semiconductor companies the functions and features that are needed to implement PLM either on the design side or on the manufacturing side. Inevitably, the need arises to customize extensively, which is an option available only in on-premise installations. This leads to several other problems, including making any potential upgrades expensive.
There are several domain-specific tools used in the overall flow and they are not well integrated. Further complicating the situation is the fact that there are several organizations/stakeholders involved in the overall process. The lack of a cohesive data ecosystem creates communication and coordination problems, leading to several problems in the product realization process. CIMData, a research company that focuses on PLM, published its findings stating that while more than two thirds of the companies they interviewed want to take advantage of being on the cloud, only 20-25% of manufacturing companies use cloud
PLM, as the current platforms don’t address their needs. They acknowledge the complexity of integrations in a cloud or hybrid ecosystem.
Research from Kalypso Partners LLC summarizes the problems plaguing the semiconductor industry as follows:
Business Situation
55% of semiconductor product launches miss original launch date
More than 40% of products exceed budget
More than 60% of designs require more than 1 re-spin
More than 47% reduction in revenue if delayed by 6 months
41% of designs don't make it to production
A root cause analysis of the same resulted in the following findings:
Root causes
Lack of reuse
Inability of engineering infrastructure to handle complex products
Lack of traceability of product evolution from requirements to manufacturing
Market pressures to shorten innovation cycles and time to market
The following table summarizes the research findings from Wilson Research Group:
| Parameter | ASIC Design | FPGA Design |
|---|---|---|
| Projects behind schedule | 66% | 70% |
| Projects behind schedule by 30% or more | 27% | Not available |
| Non-trivial bugs escaping into production | Not available | 84% |
| Percentage of project time spent in verification | 50-60% | 40-50% |
| Percentage of engineer’s time spent during verification task | 49% | 42% |
| Number of re-spins (2 or more) | 76% | Not available |
The research further attributes the functional flaws to several factors as shown in the illustration below:
Inference
It can be inferred from the above that as systems become complex, in an environment where there are several silos of applications to manage product data, it becomes imperative to ensure coordination and communication between internal and external stakeholders are effective and efficient. For this to happen, one cannot rely on manual processes. Persistence of data value chains across applications, the capability to create, easily search, visualize, and reuse IPs are key to ensure that quality of products is not compromised.
The above findings strengthen our case to provide a semiconductor focused solution that provides traceability from requirements to analysis to design to verification and validation to manufacturing. Derived from Yoke and ThoR, icSysLM will help semiconductor companies manage design blocks commonly known as IP, its lifecycle, and issues and defects related to the hierarchical design bill of IPs.
The verification process is facilitated by the accuracy and unparallel performance of icSysLM to roll up issues and defects against large SoCs and the ability to quickly see where an IP block is used. In conjunction with ThoR, it will provide customers with a powerful solution to ensure traceability from requirements to analysis, design, verification, validation, and manufacturing. On the manufacturing side, our open APIs facilitate easy integration with manufacturing execution systems (MES) and enterprise resource planning systems (ERP).
The operations bill-of-material solution will manage the process, supply chain, and qualification aspects of transforming a raw wafer to a chip. ThoR will ensure the data value chain from requirements to manufacturing is preserved. icSysLM will provide a true enterprise “Single Source of Truth” as it will promote reuse and prevent data duplication and redundancies. Companies will be able to search, visualize, and analyze connected data to make better and informed decisions. Our solution will help customers deliver projects on time, accelerate time to market, and improve operational excellence.
IC-SYSLM Overview
The capabilities of icSysLM are summarized in the illustration below:
The application is built on the fundamental premise that customers should be able to configure their way to a successful implementation as opposed to customizing it. In a typical product data management implementation, customers would like to extend the OOTB (out of the box) application data model by adding attributes, classes and relationships, behavior by way of defining custom processes, business rules and policies. Glide provides these capabilities as configuration. The goal is to empower customers to be self-sufficient and reduce the services and consulting costs by a factor of three.
Glide provides an intuitive, simple to use, user interface both on the admin side and on the user side. The platform promotes agile implementation, and it is expected that the business analyst will work with the IT administrator to configure the application. The IT administrator should have an understanding of object-oriented concepts and familiarity with BPMN 2.0 to configure the application.
Customer Engagement Model
While not all implementations are alike, with our no-code Yoke for managing domain knowledge and our low-code framework for integrating applications, we believe we will save our customers significant implementation cost. Our goal is to complete any implementation within 2 - 6 months.
Our data migration framework will help accelerate migration. The framework provides the core modules that are typically needed in a migration project. It includes standard templates for data extraction, interface adapters, error handling and resumption, data transformation modules to fit the destination system, standard loading templates for objects/relations/files, loader configuration settings, loader adapters, logging and notification cores, file transfer modules and more.
With our low-code integration framework, we can integrate with any other enterprise application like requirements management, source code control, and test management that are web-enabled and provide RestAPIs. For those legacy applications that do not provide RestAPIs or GraphQL, our professional services team will be engaged to assist with the integration.
In situations where customers already have an IP management solution and an operations BoM management solution, our application can still help integrate with other enterprise applications to ensure that the data value chain from requirements to manufacturing is not broken.
Summary
By implementing icSysLM, we are confident that customers will receive the ROI they are expecting.
- Our agile and nimble approach to defining domain specific models, processes, policies, and integrating enterprise applications will ensure that the overall costs of implementation are low.
- Services costs are very low as customers will be able to manage the implementation by themselves.
- Getting traceability all the way from business requirements, system requirements to verification and validation and manufacturing is now easier and faster to achieve and it is inexpensive.
- Integrating with project management application will provide valuable insights into project execution and resource management.
- Users get a consistent experience across the board and users find what they need faster.
- The intuitive, uncluttered user interface and its ease of use reduces the effort to onboard users, significantly reducing training costs.
- Our data migration framework will reduce the costs incurred in a typical data migration project by at least 55%.
- The integration framework will help customers integrate their enterprise applications in 2-4 weeks instead of 4-6 months. The time to value is very short with implementation in 2-6 months.
