Development of an intelligent control system for an oil refinery

Budget: 14000 EUR Deadline: 70 days

Hello! My name is Oleg, I am the project manager at Business Atlas. The described industrial AI platform is a large-scale challenge that fully aligns with our profile. We specialize in autonomous digital ecosystems and complex data processing systems. My experience focuses on low-code tools (Make, n8n) and business logic automation, while custom development (Python, PLC) is handled by our tech lead.
We fully accept your terms (NDA, GitHub, payment by deliverables) and are ready to implement the project turnkey.
Here is our technical vision and optimization of your specifications:
1. Implementation by phases and architecture
• Phase 1 (PLC / TwinCAT): Collecting signals from 42–67 points via Beckhoff CX2040, deploying an OPC UA server, configuring emergency logic and offline buffering.
• Phase 2 (Edge Server): Setting up Node-RED pipelines, databases (InfluxDB + PostgreSQL), and FastAPI. To accelerate data transfer and notification flexibility, we will integrate low-code workflows in n8n/Make locally in Docker. We will develop 5 dashboards in Grafana and a Telegram bot for auto-reports.
• Phase 3 (AI Engine): The tech lead will implement mathematical models in Python. Using scikit-learn and scipy, we will deploy gradient boosting for oil output forecasting, Isolation Forest for anomaly detection, and FFT analysis of bearing vibrations.
• Phase 4 (Frontend): Creating a React PWA for mobile monitoring of the plant with automatic generation of shift PDF reports.
2. Improvement suggestions
• Support optimization: Using n8n/Make within Docker will allow you to easily change the logic of Telegram notifications or reports in the future without involving Python developers.
• Hybrid storage: Separating hot vibration data (InfluxDB) and relational reports (PostgreSQL) will avoid server overload.
• Turnkey project estimate: from €12,000 – €18,000 (stage payments for deliverables of each phase, final estimate after diagnostics).
• Implementation timeline: 8 – 12 weeks.
Message me privately. We will sign an NDA, conduct a brief technical diagnosis, agree on the stack, and form an exact roadmap for your production! Expand

Budget: 5000 EUR Deadline: 30 days

Denis, good day.

The project sounds very interesting.

Could you please let me know if you are considering the possibility of using a different technology stack for the implementation of the project?

Budget: 5000 EUR Deadline: 7 days

Hello, I worked on a feed production monitoring system with Siemens PLC and InfluxDB for collecting data from 85 temperature and humidity sensors. We increased efficiency by 23% and reduced waste by 15%.

I’m curious if you plan to integrate the system with an existing ERP or if this will be a standalone solution for the oil plant?

I suggest we get in touch; I will provide you with free technical consultation and we can outline a development plan + I will tell you about my team! ⚙️

Budget: 12000 EUR Deadline: 100 days

Hello!

The engineering specification for your automation and predictive analytics project for the fat-and-oil production is completely clear. The architecture, combining Industrial IoT (Beckhoff / OPC UA), Edge computing, and an AI/ML engine, is designed well. At Lumvex studio, we have deep expertise in building complex automation systems, working with time-series data, and integrating AI into the enterprise sector.

We are ready to implement this project "turnkey" with our team of automation engineers (ACS TP) and Data Science specialists.

🎯 Our vision and technical contribution for key phases:
Phase 1 (PLC & Offline buffering): When programming the Beckhoff CX2040 in Structured Text (TwinCAT 3), we will pay special attention to the ring buffer mechanism on the PLC itself. In case of loss of connection with the Edge server, data from 67 measurement points will not be lost but will be synchronized via OPC UA immediately after the network is restored.

Phase 2 & 3 (Edge, FFT analysis, and AI):

Vibration (SciPy FFT): Converting the raw vibration signal of bearings from the time domain to the frequency domain (Fast Fourier Transform) is the most reliable way to catch peaks in the early stages of equipment degradation.

ML models: Training Gradient Boosting for oil output prediction and Isolation Forest for anomaly detection will occur on an isolated backend (FastAPI). We will optimize the models so that they do not overload the Edge server's processor during the continuous data flow into InfluxDB.

Phase 4 (React PWA & Expertise): We will implement a lightweight PWA application that will instantly update Grafana widgets via WebSocket/REST API, providing the factory owner with real-time monitoring from their phone even with a weak internet connection.

💼 Working conditions:
Full signing of NDA before access is granted.

Transparent workflow: commits to your GitHub from day one, task management in Jira/Notion.

Phase acceptance based on clearly defined Deliverables (submission of each phase with testing). Expand

Budget: 5000 EUR Deadline: 20 days

Good day!

I professionally develop bots in Telegram and web applications in Python. I am ready to implement your industrial AI platform with PLC integration via OPC UA, setting up an Edge server (Node-RED, FastAPI, InfluxDB), AI models for predictive analytics (scikit-learn), and React PWA + deploy the entire infrastructure in Docker.

Message me privately, and we will clarify the details.

Budget: 4800 EUR Deadline: 5 days

Hello!
This is a serious industrial project — and I am interested in it. I have the technical foundation to implement most phases, but I will be honest about each one.
What I can confidently cover:
Phase 2 — Edge Server: Node-RED, InfluxDB, PostgreSQL, FastAPI with JWT, Docker-compose, Telegram bot — this is my main stack, I will implement it without questions.
Phase 3 — AI Engine: gradient boosting for oil output prediction, Isolation Forest for anomalies, FFT vibration analysis via scipy, predictive maintenance — I know the Python ML stack well, I have experience building production ML pipelines.
Phase 4 — React PWA, PDF report generation — I will implement.
Where clarification is needed:
Phase 1 — TwinCAT 3 / Beckhoff PLC programming — this is a narrow industrial specialization. I will connect the OPC UA server and signal collection from the Edge side, but Structured Text programming of the PLC is better handled by a specialist with experience specifically in TwinCAT. If you already have a PLC specialist — I will take care of the rest. If not — I can involve a partner for this part.
Approach to work:
GitHub from day one, phased deliverables, I will sign an NDA. I can provide a technical specification for each phase before starting so you understand exactly what you will receive.
Estimated cost (excluding Phase 1 PLC):
Phase 2 — $1,500–2,000
Phase 3 — $1,800–2,500
Phase 4 — $800–1,200
I am ready for detailed discussion. Write to me — we will schedule a call.
Best regards,
Bohdan Expand

Budget: 10000 EUR Deadline: 100 days

I worked on https://poseidon.codezerogroup.com - an enterprise Python/React platform with real-time monitoring, JWT, dashboards, and data pipelines. Your Edge + AI + Frontend architecture is the same scope, just in the industrial domain.

Industrial AI systems for production monitoring are my core. Full stack: from hardware layer through data pipelines, time-series databases, APIs, to ML and operator dashboards.

What I will do:
- Edge Server: Node-RED, InfluxDB, PostgreSQL, Docker-compose
- API FastAPI + JWT, 5 Grafana dashboards, Telegram bot with reports
- AI Engine: gradient boosting, vibration FFT, Isolation Forest, predictive maintenance, ROI report
- Frontend React PWA, PDF reports with changes
- (Option B/C) Beckhoff TwinCAT 3 PLC, OPC UA, 42-67 points, emergency logic

--- OPTIONS ---

- Option A (Core - Phases 2-6, without PLC): 10,000 EUR (100 days) - Edge Server, AI Engine, React PWA. Without PLC.
- Option B (Full Stack with PLC): 16,000 EUR (160 days) - everything from A + Beckhoff TwinCAT 3 PLC programming, OPC UA, emergency logic. - best scope/price ratio
- Option C (UVWeb Premium - 6 months SLA): 20,800 EUR (180 days) - B + 6 months support, retraining AI models, SLA.

Execution time: 100-180 days. I need specifications for measurement points, PLC network diagram, and GitHub access from day one.

Portfolio:

- https://poseidon.codezerogroup.com - enterprise monitoring, Python/React, JWT, dashboards
- https://nyxvx.com - Epic Games, custom CMS, high-load infrastructure
- https://ou-uv.com - Flask/Python CMS, multilingual, API, Docker

8 years in backend and AI - from simple APIs to industrial platforms with ML, IoT, and monitoring.

Available immediately, I can discuss the architecture and start this week. Expand

Budget: 5000 EUR Deadline: 30 days

Hello!

Thank you for the detailed project description.

After preliminary analysis, we believe that we can take on most of the work for this project, including:

• Backend development (FastAPI, PostgreSQL, InfluxDB)
• Data collection and processing via Node-RED
• Development of a React PWA interface
• Grafana dashboards and analytical panels
• Telegram bot and automated reporting
• Docker infrastructure and DevOps setup
• AI/ML modules for forecasting, anomaly detection, and production process analytics

Based on our preliminary assessment, we can cover about 80–90% of the stated workload.

Before preparing the final technical and commercial proposal, we would like to clarify a few important points:

1. PLC / Beckhoff TwinCAT

• Is there a need to develop or modify the PLC logic on Beckhoff TwinCAT?
• Or are the PLC and OPC UA already functioning, and the task is only to obtain and process data?

2. Historical data

• What volume of historical data is currently available?
• Are there data on production, oil output, maintenance, accidents, downtimes, and equipment failures?

3. Sensor readiness

• Are all measurement points (42–67) already installed, configured, and calibrated?
• Or does the selection, installation, and integration of sensors also fall within the project scope?

4. Deployment environment

• Is a fully local (on-premise) deployment planned, or is the use of cloud infrastructure allowed?
• Are there cybersecurity requirements or restrictions on connecting to external networks?

5. Expected results of AI modules

• What KPIs or performance indicators are expected for forecasting oil output, anomaly detection, and predictive maintenance?
• Are there target accuracy metrics for the models?

6. Commissioning work

• Is it necessary for specialists to visit the site for system implementation and launch?

7. Interaction with enterprise experts

• Will there be an opportunity to work with engineers, technologists, and maintenance personnel of the enterprise to clarify operating modes, accident thresholds, and equipment specifics?

After receiving this information, we will be able to more accurately assess the workload, implementation timelines, project budget, and propose an optimal solution architecture.

We would be happy to discuss the project in more detail. Expand

Budget: 5000 EUR Deadline: 10 days

I have done projects of this kind! We will do it for you too! Many such cases.

Budget: 5000 EUR Deadline: 10 days

Hello.

This is a very simple task for me.

Please contact me as soon as possible.

Let's discuss this together.

Budget: 4990 EUR Deadline: 29 days

Good day! My name is Valentin, and I represent Arctic Web Agency. We are a team that specializes in creating modern and effective solutions for businesses. I can provide examples of our similar work in personal messages. We are ready to take your project to work!

Sincerely,
Arctic Web Team
Freelancehunt

Budget: 5000 EUR Deadline: 30 days

Hello!

I am a Full-Stack Software Engineer with over 7 years of experience in developing websites, SaaS solutions, complex web platforms, and MVPs for startups - from idea and architecture to production and support.

I work not only as a developer but also with a focus on business logic, scalability, and long-term support of solutions. My portfolio includes examples of implemented projects of varying complexity.

Technology stack:
PHP (Laravel, Symfony, Yii2),
Frontend: JavaScript (Vue.js, React.js), HTML5, CSS3,
Databases: MySQL, PostgreSQL.

Regarding timelines and costs - at this stage, it is difficult to accurately assess them without clarifying the requirements. To form a reasoned estimate, it is usually necessary to understand:
- project goals;
- key functionality;
- expected load and integrations;
- priorities (quick launch vs scalability).

I suggest starting with a brief clarification of the requirements or a call / correspondence - after that, I will be able to provide a realistic estimate on timelines, budget, and implementation options.

I would be happy to discuss the details. Expand

Budget: 6000 EUR Deadline: 60 days

Hello!

This sounds like exactly the type of industrial platform where small details really matter, especially when it comes to security logic, data integrity, and preventing signal loss during unstable network connections. I have created edge solutions using time series storage, dashboards, APIs, and ML modules, and I am comfortable working directly with hardware signals while maintaining a clean and well-tested software architecture.

Regarding PLC, I would consider the CX2040 as the primary source of truth for all safety and emergency behavior scenarios, and on top of that, I would build a robust OPC UA model with clear naming and units of measurement, as well as local data buffering to avoid gaps during network failures. On the edge server, I can set up easily maintainable Node-RED flows, store raw and processed data separately in InfluxDB and Postgres, and provide access to everything through FastAPI with JWT authentication and a role-based access model so that operators and owners see exactly the level of information they need.

As for the AI part, I like your direction. As an improvement, I would suggest adding a simple model monitoring loop to track the quality of predictions over time and allow the system to notify when the model starts to degrade or lose accuracy. My key idea is a single Factory Health Score metric that updates every minute and is explained in simple language directly in a Telegram bot: a short message "why" and a list of the main sensors that had the most impact on the score change. Owners like this because the information is easy to understand, but they can always drill down into detailed graphs in Grafana.

I am completely fine with NDA, transferring all rights to the client, working through GitHub from day one, and a phased implementation of the project with careful handover of results. If we talk about a more reliable implementation option, I would also suggest running all components in containers but isolating the data layer from the very beginning by setting up backup and storage policies, as fixing such things at later stages usually becomes very expensive.

https://live.chatbullet.com
https://oscarstories.com

Thank you! Expand

Budget: 5000 EUR Deadline: 30 days

Good day, I represent the company Sunpoint IT Solution. We have been in the IT market for over 9 years. We have a serious team of specialists and we would be happy to take on your project. Website https://sunpoint.com.ua

Budget: 5000 EUR Deadline: 20 days

Good day. I am ready to complete this project as I have extensive experience in application development.

Budget: 5000 EUR Deadline: 11 days

Dear Denys

I hope this message finds you well. I have carefully reviewed your project requirements and fully understand the scope, including PLC programming, edge server setup, AI analytics, dashboards, and the mobile PWA.

With extensive experience in Beckhoff TwinCAT 3, Python, Node-RED, InfluxDB, PostgreSQL, FastAPI, Grafana, Docker, and React, I have successfully delivered industrial IoT and predictive maintenance solutions. My expertise also includes AI-based yield prediction, anomaly detection, vibration analysis, and automated reporting.

By partnering with me, your company will gain:

Optimized oil production and reduced downtime
Predictive maintenance to prevent failures and save costs
Real time dashboards and automated reports for actionable insights
Scalable, maintainable, and fully documented solutions
Full code ownership under NDA

I would also recommend considering MQTT to improve scalability and decoupling between PLC, edge, and AI layers. I am ready to discuss milestones, deliverables, and turnkey implementation options to meet your goals.

Thank you for your time and consideration. I look forward to the opportunity to contribute to your project’s success.

Warm regards,

I can also introduce my powerful team.
I believe that by working together, we can create miracles.
I look forward to your reply.

If you want, I can also create an even more engaging version that subtly emphasizes reliability, trustworthiness, and speed qualities clients love when hiring for industrial automation projects. Do you want me to do that? Expand

Budget: 5000 EUR Deadline: 28 days

Hello!

Our team is ready to implement the project "turnkey" — from design to full development and launch. We have experienced developers, designers, and UX/UI specialists in our team, so we can create a convenient, modern, and functional product according to your requirements.

We will be able to determine the deadlines and budget accurately after discussing the details and understanding the full scope of work. I would appreciate your feedback; I can send examples of our work in private messages.

Budget: 5000 EUR Deadline: 25 days

❤️Hi Denys, hope you are doing well

I read the full scope and understand you need an industrial AI platform for sunflower oil production monitoring, covering Beckhoff TwinCAT 3, Python 3.11, Node-RED, InfluxDB 2, PostgreSQL, FastAPI, Grafana, Docker, React PWA, Telegram Bot API, scikit-learn, and scipy FFT.

I can help structure this in clear phases: PLC/OPC UA signal collection and emergency logic, edge pipelines with InfluxDB/PostgreSQL, FastAPI JWT APIs, Grafana dashboards, Telegram reporting, AI models for oil output prediction, vibration FFT, Isolation Forest anomaly detection, predictive maintenance, and a mobile React PWA.

My focus would be reliable industrial architecture, offline buffering, secure APIs, maintainable Docker infrastructure, GitHub workflow, documentation, and phased deliverables under NDA.

I can provide an accurate timeline and cost after reviewing plant signals and required dashboards.

Best regards,
Youssef. Expand

Budget: 5000 EUR Deadline: 30 days

Good day, I have experience in Beckhoff programming, installation and implementation of monitoring on Grafana on the same Beckhoff, Node-RED without problems, I can send project samples + I do visualization on Beckhoff (I hope you won't need to do this). I will answer all your questions. My personal opinion - better to use Schneider or Siemens PLC + PC with Grafana, than to struggle with the TwinCAT development environment. If this is not far from Barcelona, there are "remote hands" available to connect the hardware.

Budget: 25000 EUR Deadline: 75 days

We already have a practically ready similar solution for industrial monitoring with AI analytics that can be quickly adapted to your plant and discussed here on the marketplace)).
Regarding the budget - 5000 EUR is realistic for technical design or the first stage, and for a turnkey solution, I would estimate starting from 35000 EUR and 60-90 days, depending on the readiness of signals, PLC logic, and safety requirements.
NDA is fine, GitHub from day one is fine, the code and documentation remain the property of the client.

I see the implementation in stages - first architecture and signal mapping, then PLC and OPC UA, next an edge server with InfluxDB, PostgreSQL, FastAPI, Grafana, and Telegram reports, then AI modules for predicting oil output, FFT for vibration and anomalies, and finally React PWA and PDF reports of changes.
We can keep it simple - it's better to launch AI after stable historical data collection; otherwise, the model will be more speculative than helpful, and that's already philosophy in a bad sense =/

> Is there already an exact list of 42-67 measurement points with sensor types, polling frequency, and alarm limits?
> Is the Beckhoff CX2040 already installed in production, or does the PLC part also need to be designed from scratch?

Relevant examples from Ingello:
> https://business.ingello.com/vorfahr - AI automation and complex decision-making logic
> https://business.ingello.com/forma-bpm - process management systems, roles, panels, operational logic
> https://systems-fl.ingello.com/ua - our main profile for such tasks on the marketplace

I am available here; we can agree on the first deliverable - technical project, architecture, risk plan, and accurate cost estimate before development. Expand

Budget: 5000 EUR Deadline: 15 days

Hello!

The industrial contour OT + edge + AI + PWA is an interesting project, the stack is logical.

Phases:

TwinCAT CX2040, 42-67 points, OPC UA, accidents, offline buffer
Node-RED → InfluxDB + PostgreSQL, FastAPI JWT, Grafana (5 screens), Telegram, Docker
ML: oil output prediction, FFT vibration, Isolation Forest, PdM, ROI report
React PWA + PDF changes
Conditions: NDA, GitHub day 1, code - client, payment by deliverables.

Turnkey (orientation): €30-50k or by phases (1: €8-14k, 2: €10-16k, 3: €7-12k, 4: €5-8k). Can do 1+2 without AI, then ML on the data.

Team: TwinCAT + Python/ML + React.

From you: list of sensors, access to PLC/stand, NDA.

Ready to discuss scope and price after NDA. Expand

Budget: 10000 EUR Deadline: 60 days

Hello, Denis!

I am engaged in the development of such systems. The latest project is a production management system for plastic products.

I suggest we discuss the project in more detail in private messages. We will sign all necessary contracts. We will do it qualitatively, quickly, and within a reasonable budget.

It is currently difficult to estimate the cost realistically, as you have described the functional requirements superficially, and the non-functional ones need to be discussed. I would estimate a range from 10,000 to 30,000 euros. But this is still more of a shot in the dark.

Budget: 5000 EUR Deadline: 40 days

Hello! Ready to collaborate. I offer quality and fast work. I have experience in creating automated systems. Write to me.

Budget: 5000 EUR Deadline: 1 day

Good day.. I have experience in structuring complex projects and I have questions regarding the architecture of implementation and interaction between individual blocks of the system. For example, who is responsible for integration between phases, how the results from the AI block will be accepted, who owns the rights to the code, models, and data (IP), governance of GitHub repositories and access, and how the transfer of the system will be organized after the work is completed. If this is relevant to your project, I would be happy to discuss the details and cost.

Budget: 5000 EUR Deadline: 31 days

Hello. I was interested to read about your project. I am confident that I can deliver effective and high-quality work that meets your requirements and expectations. I have over 8 years of experience. I am ready to discuss the details and start working. I will be waiting for your response, feel free to write and we can discuss.

Budget: 5000 EUR Deadline: 22 days

Hello, I have extensive experience working with both low-level industrial hardware (PLC, Modbus/OPC UA protocols, control systems) and building high-load Edge/Cloud platforms (FastAPI, Docker, InfluxDB) as well as developing ML models. Your architecture looks solid and well thought out, with a stack tailored to high industrial standards.
Below are the phases of work that I see:
Phase 1: PLC & OPC UA
Programming Beckhoff CX2040 on TwinCAT 3 (ST).
Collecting signals from all points (up to 67 sensors).
Setting up the OPC UA server and offline buffering logic.

Phase 2: Edge Infrastructure & Storage
Deploying Docker-compose (FastAPI, InfluxDB for metrics, PostgreSQL for relational data).
Setting up data collection pipelines, 5 screens in Grafana, and a Telegram bot for reports.

Phase 3: AI Engine & Predictive Analytics
Developing a model to predict oil output (Gradient Boosting).
Implementing FFT analysis of bearing vibration frequencies (scipy.fft).
Launching an anomaly detector (Isolation Forest) and ROI calculation algorithm.

Phase 4: Frontend PWA & Reports
Developing a mobile React PWA application for the owner.
Automatic generation of PDF change reports (via WeasyPrint).
Comprehensive load testing of the pipeline and project handover.

I would be happy to work together; my GitHub is https://github.com/onyx144 Expand

Budget: 5000 EUR Deadline: 1 day

Hello. I am ready to deliver a turnkey solution, guaranteeing quality and promptness. I am willing to work with an NDA, via Git, and discuss all the main points. I am very interested in the project, the stack is clear.

Feel free to message me privately, I would be happy to collaborate!

Budget: 5000 EUR Deadline: 3 days

Hello. Are you planning to use existing machine learning models for predicting oil output, or do you need a complete development from scratch based on your historical data?

I will clarify the details regarding timelines and budget in personal correspondence.

Here’s how I will implement this project:
1. In the PLC phase, I will program TwinCAT 3 with signal collection, emergency logic, and an OPC UA server.
2. In the Edge Server phase, I will set up Node-RED, InfluxDB, PostgreSQL, FastAPI with JWT, Grafana, and a Telegram bot in Docker.
3. In the AI Engine phase, I will deploy a gradient boosting model, FFT analysis, and an anomaly detector for predictive maintenance.

Thank you for considering my proposal. I look forward to the opportunity to collaborate with you!

Budget: 5000 EUR Deadline: 3 days

Hello. I can evaluate and execute this as a separate MVP, ideally in Python, unless there is a requirement for a different stack.

Before starting, we need to finalize the exact list of features, the format of the result, and the criteria for completion. The cost includes the implementation of the agreed specifications and one round of minor revisions. New features or changes in logic after the start will be evaluated separately.