1. What is Calejo Optimize?

Calejo Optimize is an optimization engine for industrial processes. It represents a whole new way of modelling and helping the industry to manage, structure, analyze, simulate, optimize and control the overall data and knowledge flows from its processes. Calejo Optimize has been developed based on the latest scientific findings and represents the unique result of 15+ years of experience from various international AI projects.

2. How does Calejo Optimize work?

Together with the customer, we collect, wash and analyze important input data from the process, after which we create a so-called digital twin. Our digital twin is a hybrid solution, which can be based on both traditional physical calculation and what we usually call AI, that is, the connection between data.

With the help of the digital twin, real-time data from the process control system and production planning, the operator can get suggestions on how process parameters should be adjusted to achieve an optimal and sustainable total production.

The solution can most easily be compared to the type of navigation system used in cars. The navigation system relates to existing conditions and existing traffic or process environment. But it can also take into account external and more varied factors and temporary deviations, such as maintenance work, queues and disturbances in connection with the car or as in this case the process navigator’s route suggestions. It can also suggest an update of the route during the journey.

In the next step, the digital twin can also be used to train, simulate, optimize and ultimately also control an autonomous industry.

Calejo Optimize 1.0 is based on Open Modelica. The different parts of the model are linked to the database. In the library there are unique AI blocks to be used. The result is presented in existing program dashboards, SCADA systems etc. or in a separate instrument panel, which is part of Calejo Optimize. The system can collect data from all modern platforms and systems. Ie IoTplatforms, Azure, Siemens, ABB, SAP, SQL, Oracle, Maximo, etc.

3. How do you work with
your customer companies?

During a joint workshop on-site with the customer, the functions and data flows of the current process are mapped to obtain consensus around boundaries and what effect goals the use of Calejo Optimize will achieve.

Based on set goals, Calejo’s representatives then announce the conditions required in terms of the amount of existing data, the handling of real-time data and the configuration of the control system.

Implementation takes place in the following stages.

Stage 1. The creation of a digital twin.

– Process mapping.

– Data verification based on measuring points, quality and quantity.

– Modelling a digital twin over the process.

– Training the digital twin.

– Verification of the digital twin against the actual process.

Stage 2. Optimization of the process based on given power goals.

– Priorities and rules for optimization are developed. Calejo Optimize can optimize processes based on more than one effect target. A mutual priority between the goals must thus be created. Other process critical boundaries and rules must also be defined. The system is optimized on the basis of given conditions.

Stage 3. Interface for operators.

– Together with selected operators, the interface between Calejo Optimize and the operators is adapted so that the instructions’ designations and nomenclature correspond with the other operator environment.

Stage 4. Verification and final delivery of the implementation.

– Operator support is being rigorously tested by Calejo’s staff to verify the implementation.

– Training of operators and others concerned with the use of Calejo Optimize.

– Commissioning of Calejo Optimize.

4. What does we as a customer ourselves need to set up in terms of resources, expertise and commitment?

The customer determines the degree of commitment. It is the customers who is the experts in their own process. We discuss together the best solution and what is feasible based on the conditions that exist.

What Calejo needs is input from the customer about how the process works and also help with collecting data. If the customer does not have their own knowledge or can provide their own resources, Calejo can help with guiding, or contribute with resources from one of Calejo’s many partner companies.

Calejo Optmize does not require any special hardware solution. The system runs on any hardware – CPUs, GPUs, TPUs, Linux, Windows…

5. How do you manage
our company if:

A. We believe that we are not mature enough today to implement your solution. We have not progressed far enough in our digitalization. How can we still go about utilizing your technology and expertise?

Calejo works closely with several of the market’s leading technology consulting companies. Together with these, we can help with most things when it comes to supporting your company’s continued digitization process. This includes the implementation of IoT, AI training of management groups and operators, control and verification of data flows and the creation of a digital twin over the plant’s processes. The customer adds competence to their operations and processes and determines the roadmap.

B. We believe we are at the forefront of our digitalization and adaptation to new technology and AI. What can you add that we don’t already have?

What makes Calejo unique is that we have developed a solution with which we can model and work with very complex processes and many different parameters. With the help of Calejo Optimize we can combine mathematical models with AI modelling in our own developed and patented system that simultaneously and instantaneously can handle very complex systems faster and with considerably larger amounts of data than other available offers on the market. See also the answer under point 11.

6. Where do you get the best results – when designing new plants, in ongoing production and/or in maintenance?

Our technology works in all these processes. When designing new plants we can

build a new mathematical white box model that we then continuously can improve. In continuous production, our solution is given, and in connection with maintenance we can use our models to see how all or part of the processes change over time.

7. What conditions are needed for Calejo Optimize to be implemented and used?

We must have data, unless data not is available we must find ways to collect data. However, no much data or long collection time is required. Under normal conditions, historical data from three months back in time is needed depending on the process and how many data points are collected per unit of time.

We can install our solution with aggregated data from the process locally inside the facility or in private or public clouds. We can also – if necessary – with help from our partners create appropriate systems and cloud solutions to collect and store all information.

When it comes to control systems, we are not dependent on any special ones. Calejo delivers an overall control system, so it does not matter which control system the customer has chosen to use. Calejo already cooperates or can cooperate with most control systems on the market.

The customer also does not need any AI skills. However, together with our partners, we are happy to assist and train operators and management to ensure that employees have a better understanding of the possibilities of technology and increase their own and the company’s overall AI maturity,

However, what the customer – or a customer-close consultant – must contribute is business understanding of their processes. This is so that we can jointly and as accurately as possible determine what is to be achieved.

8. We have already built a digital twin of our process. Can we use it?

Yes, there are other companies that can help create digital twins of processes. Most of these digital twins are mathematically structured to help companies gain a better understanding of the process. What makes Calejo unique is that we can combine these mathematical models with our AI modelling in our own developed and patented system that can simultaneously and momentarily handle significantly larger amounts of data than other available offers on the market.

9. What references do you have from previously completed and successful projects?

We have done a successful PoC with Uniper within hydro power, which confirms that Calejo Optimize can be used to better understand and optimize production. The Hydro Power system consist of several reservoirs and generators.

During 2019 we have started successful projects with SCA, ABB, Boliden, Optimation and Processum. We are also planning for projects with several of Sweden’s most AI-mature companies and we have started collaborative discussions with some of the largest and leading technology consulting companies in Sweden, all of whom are very interested in being able to collaborate around Calejo Optimize.

Together with SCA Calejo optimizes the energy within the pulp process in SCA Obbola by forecasting the need for steam and using accumulated energy within the process to ensure that there will always be enough steam.

Together with the mining company Boliden Mineral AB and the process optimization company Optimation Calejo are developing a new method for digital twins of complex processes using AI.

In 2020, together with SCA and ABB, Calejo will also create operator support applied in a control room at SCA Östrand’s pulp mill. The support will be in the form of visual feedback of the process’s future development of selected process values given the current parameter setting as well as suggestions on how and when the operator should adjust selected parameters to get an improved process outcome.

The solution will mean that the process can be optimized towards more than one goal. Among other things, the optimization will be able to give the operator instructions that lead to reduced energy consumption and the use of harmful substances, while maintaining quality and reliability. Thanks to ABB’s involvement in the project, the visualization and operator instructions will be presented in the control room’s existing operator displays.

10. What goals can your solution help us achieve?

We can help you achieve many different goals. First, however, in a joint workshop we must decide which of all these goals are the most important priority. We can then help to optimize the process based on several of these goals and weight them among themselves.

The immediate benefits simulations and optimizations through our solution can contribute with is a better understanding of the processes and the ability to secure the skills and the transfer of skills. Other benefits include increased efficiency, increased operational reliability, increased productivity and availability, increased sustainability and the opportunity to achieve the environmental goals, as well as improved quality.

11. What is unique and
new about Calejo’s offer?

Calejo’s uniqueness rests in our patent-pending solution with instant gray box modelling that can include very disruptive data sets. Grey box modelling or hybrid modelling is a combination of the two modelling languages and techniques – purely mathematical white box modelling and AI black boxing. In our solution we can handle and derive all calculations in the whole process as a single large model. Our optimization engine can handle a large amount of data faster than any other existing system.

Uniqueness of Calejo Optimize:

– Patent pending solution:

– Non-linear modells.

– DAE compatible and faster scalable than other solutions.

– Variable time-step solving.

– Faster optimization than reinforcement learning.

– Easily deployable in edge devices, cloud and personal computers.

– Mixed neural network and physical equations in any combination.

12. Can Calejo Optimize be
used to map and control all processes?

Yes, our Calejo Optimize solution is generic and can be applied to most processes. The more complex the process is, the harder it is to predict and the more benefits you have of our offering. We have therefore chosen to begin by approaching the process industry, as it is complex and collects a lot of data from the beginning. Here, the quality of the process is also dependent on having good control over temperatures, flows and other important parameters. The processes are also usually repeatable and independent of human impact.

13. How is Calejo’s
offer packaged?

The implementation phase:

We always start our customer relations with a workshop, which is free of charge if the customer chooses to proceed with the construction of a digital twin.

Once we have agreed with the client on the scope of the project, we quote a cost for implementation and training.

The subscription offer

We then sell Calejo Optimize as a subscription service with a monthly subscription cost depending on the process and size.

A subscription to Calejo Optimize includes the following.

– Access to the implemented AI support for increased process efficiency and reduced environmental impact.

– Support from Calejo service personnel.

– Access to updates and enhancements in the underlying software.

– Calejo Optimize is self-learning and can therefore automatically adapt to process changes over time without external intervention.

The following services are available as an option.

– Warnings to maintenance personnel on changes in the process that could lead to future operational disruptions.

– Ability to use Calejo Optimize for training of operators.

– Ability to use Calejo Optimize for better process understanding.

– Optimization of production taking into account future price trends.

– Optimization taking into account other external parameters, such as commodity supply forecasts, currency changes and energy costs.

14. What does Calejo Optimize cost?

Although the system is developed for the process industry, it still needs to be adapted to the customer’s own processes. Depending on complexity and scope, the need for customization varies. This means that also the price for implementation and subsequent subscription costs will vary. An extensive process entails increased costs, but also has a greater optimization potential, thus increasing Calejo Optimizes ROI with increased complexity.

– The costs of the initial workshop at 5,000 Euro is charged only if the customer does not complete the implementation.

– The costs for implementation and training are quoted before and are entirely dependent on the scope of the project.

– The subscription to Calejo Optimize is charged per month depending on the chosen scope of service.