FREQUENTLY ASKED QUESTIONS
Connected Factory Solutions are designed for plant owners to transform manufacturing operations across their enterprise. Connected Equipment Solutions are designed for equipment or product owners to monitor and optimize their fleet of equipment deployed within end-user customer's enterprises.
TwinThread optimizes key metrics related to quality, production, uptime, energy efficiency, asset reliability, and asset life. Our pre-built solutions target specific sets of metrics for improvement. For example, our Predictive Quality solution targets metrics like First Pass Quality, and Run CpK and our Predictive Energy Efficiency solution targets Energy Cost / Unit production.
TwinThread solves for key problems across both operations and maintenance by applying a modern, repeatable, highly digitized, rule based, and data-rich approach to Continuous Improvement that is deployed quickly and can be used by everyone in your organization. TwinThread has overcome many of the technical barriers to adopting advanced operations technologies, including connectivity to virtually any vendor's sensors, automation systems, and industrial software coupled with a unique ability to scale solutions across many hundreds of thousands of assets.
TwinThread connects to virtually any Historian, MES, CMM, DCS, PLC, Sensor Gateway or IoT platform in minutes. Our Edge Agent technology performs local processing and maintains a secure connection to the TwinThread Cloud. TwinThread's predictive applications can be configured in less than an hour to proactively drive improvements across key operations and maintenance metrics.
Click here if you'd like to learn more about TwinThread's architecture.
The TwinThread team has unmatched experience and a 30 year track record of innovation in industrial software. That experience translates to speed of implementation, scalability of deployments, and ultimately high ROI for your organization.
To learn more, view our about page.
TwinThread is the only Predictive Operations vendor that offers a simple and transparent pricing model - providing quotes over the web. It's important to us that your price doesn't change based on the size of your problem and that your adoption of our technology is rewarded in a transparent way. The more you deploy, the lower your unit costs - accelerating your return on investment. Our pricing is based on a simple count of "sensors" or inputs being fed into the system, as well as the number of inputs you wish to analyze or model. This approach allows you to start small and scale to the enterprise level - paying only for what you need, down to single "sensor" resolution.
View our complete pricing page here.
Different solutions require different amounts of data. First principal based metrics and models do not require any history, simple anomaly models can deliver value with a few weeks of history, and more complex solutions such as Predictive Quality may require up to a year's worth of data to become effective. The good news is that if you have historical data already its very likely we can tap directly into it. if you don't have historical data there is no better time than the present to start collecting that history to begin unlocking the potential of that data.
TwinThread's predictive applications each offer two categories of value. The first category is "learning" based value - this comes from training models and discovering what they have learned about your process / equipment and taking action based on those learnings. An example of this would be using a Predictive Quality model to understand what process recipe should be used to achieve best product quality. Typically, these kinds of learnings get applied through an existing process of managing operator guidelines or recipes / limits within an automation system. The second category of value comes from "recommendations" - these are real-time recommendations provided to your operations team to improve performance immediately. Sticking with the Predictive Quality example, if a quality parameter is off target, TwinThread provides the real-time feedback necessary to bring quality back to target in the most efficient way. Applying both categories of value is essential to driving sustained improvements in performance.
Deployments typically enjoy a 10x return on subscription investment and can be implemented and fully operationalized in 4-8 weeks. The highest continuous ROI applications are Predictive Quality, Predictive Uptime, and Predictive Throughput. These predictive applications can have an immediate and continuous impact on performance, achieving payback periods as short as 11 days (for a full year's subscription). Predictive Energy Efficiency also offers continuous return, but typically has slightly longer payback periods of approximately 30 days. Asset Reliability and Asset Life offer "event" driven and cost avoidance ROI profiles (making it more difficult to accurately approximate a continuous ROI). However, preventing a single catastrophic failure often justifies the price of a subscription many years over.
Read TwinThread's success stories here.
TwinThread offers multiple implementation options, from completely self-provisioned to fully turnkey. As part of our base subscription, we offer a host of onboarding resources including online support, regularly scheduled Onboarding Accelerator webinars, and no-agenda office hours where you can interact live with TwinThread's Customer Success Consultants. These resources are provided to accelerate your adoption and ensure your self-sufficiency in managing the platform. The most popular implementation option is our "assisted" onboarding, where TwinThread engineers (or our partner engineers) work directly with your team to implement the first use case on the first process / asset with the goal of supporting you in achieving self-sufficiency within the first implementation phase. If you feel you don't have the necessary resources for either of the implementation approaches mentioned above, TwinThread (and its partners) can offer turnkey onboarding as well as a fully managed service deployment.
TwinThread can be fully operationalized in 4-8 weeks. The software itself can be deployed in minutes to hours, while the remaining onboarding time is focused on how to take learnings and recommendations coming from TwinThread's predictive applications and incorporate those into your operating strategy.
Typically, there are three types of systems that TwinThread connects to and works with. On-Premise Automation and Information Systems are the first type. This includes PLCs, DCSs, Historians, and MES systems. TwinThread has developed deep integrations with the most popular vendors' solutions. IoT Gateways and Smart Sensors represent the second type of system TwinThread works with. Our "IoT Hub" enables direct streaming data connections from a variety of vendors' sensors, using multiple protocols, and over a variety of communication networks. The third and final connection type is Cloud-to-Cloud application connections, including leading cloud vendors and industrial cloud applications. Similar to On-Premise, we have developed deep integrations with the most popular vendors' solutions.
TwinThread has an On-Premise Agent that connects to local automation and securely pushes data to our Cloud. We also make use of our "IoT Hub", which allows you to create your own service (running locally on your device) and securely push data to TwinThread's Cloud. In both cases, secure HTTPS is used and the Agent can still securely operate where corporate firewalls are required to block inbound traffic.
TwinThread is built from 30 years experience optimizing thousands of plants and millions of assets across virtually every industry. We have applied TwinThread to virtually every process type from continuous, to batch, to semi-continuos, to discrete repetitive. Where we really shine is when you need a solution that seamlessly handles multiple process types and/or multiple types of equipment. This capability comes from the diversity of experience of the team and our history of innovation in industrial software across many industries.
To learn more, visit our about page.
Our Customer Success team brings a wealth of industrial experience with backgrounds in control systems engineering, process engineering, industrial consulting, and IT. As the team's name indicates, their mission is your success - whether you are self-provisioning your first solution or you are rolling out across your enterprise, this team has seen it all and is here to support you every step of the way. TwinThread also has a stable of highly experienced partners who are experts in bringing together OT and IT to transform your operations.
The first key step is to identify a problem worth solving. Pick a metric that has clear room for improvement, work with the TwinThread team to identify the solution that improves that metric, and execute a production pilot on a limited number of assets / lines to impact that metric. Then, use the success of that production pilot to justify and fund a roll-out. Provided you follow our Onboarding Accelerator process, we make it very easy and low risk for you to execute a production pilot and build your justification for roll-out in 4-8 weeks. We are not interested in working on science experiments. We want to move the needle for your business and deliver significant ROI.
If you've identified a problem that is worth solving and a metric you'd like to optimize, contact us today. We'll start by briefing your team on our predictive technology and onboarding approach. Then, we'll follow that up with an opportunity assessment with one of our Industrial Consultants. After that, it's just a matter of identifying your pilot team and executive sponsorship prior to kicking off your production pilot.
Security and trust are our highest priorities. We understand it is a privilege to work with your data and an imperative to protect it. You own your data as well as the learnings that TwinThread reveals from that data. Under no circumstances will we sell your data in any form (anonymized or otherwise). We will protect your data by applying best security practices within the Microsoft Azure Cloud platform, committing to building secure software from the ground up, and evolving our security controls to address continuously evolving threats. TwinThread is routinely audited by some of the world's largest companies and we encourage all customers to actively engage with us to keep their data secure.
Yes. TwinThread offers an "add-in" for Jupyter Notebook that provides a tight integration with our Model Factory. This allows a data scientist to take full advantage of TwinThread's data management and contextualization infrastructure, empowering them to focus their full attention on developing better algorithms. Custom-developed algorithms can be set up to "compete" against TwinThread's out-of-the-box algorithms in the platform. TwinThread will automatically apply the algorithm that achieves the best results in each specific application of a model. Developers may also use the Model Factory to build their own complete models using powerful building blocks that TwinThread provides.
Learn more about TwinThread's Key Capabilities.
Absolutely. Process and equipment experts are "first-class citizen" users within TwinThread. Frankly, we built this product for them. We want them to have the ability to innovate quickly using data without having to be an IT or data science expert. All they need is their process and equipment expertise - TwinThread makes everything else easy.
We install a lightweight piece of software called an Agent which communicates with your data sources (whether they be Historians, PLCs, or other data sources) and sends key data up to the TwinThread cloud for analysis.
A Digital Twin is a digital representation of your assets and processes organized into an easy to understand and visualize format. A Digital Thread connects to a Digital Twin to create calculations, rules, and actions that assess and improve performance.
There is no way to assess the predictive power of your data without configuring a predictive solution, training that solution based on your data, and reviewing model results. The good news is that TwinThread's out-of-the-box templates allow you to go from data connection to complete application in less than one hour. This approach allows you to experiment, evolve, and innovate to discover the predictive power of your data faster than any other method.
There are two key steps to solving performance issues. The first is to find the root cause of the problem. The second is to ensure the resolution to that problem is consistently applied. Making sure problems don't recur and performance does not degrade over time is a critical function of your Predictive Operations platform.
The entire feature set of TwinThread is available with your subscription. Our pricing is based on a simple count of "sensors" or inputs being fed into the system, as well as the number of inputs you wish to analyze or model. This approach allows you to try any of the predictive applications and only pay for what you use.
More information about pricing can be found here.
The only hardware that may be required for TwinThread is a workstation or server to run our Agent software. This typically is a very small footprint Windows computer or virtual machine.
It is not unusual to connect to your data sources in a matter of minutes and to see your first results in a matter of hours.
Our QuickStart wizards make building out your Digital Twins effortless.