Distributed Infrastructure Monitoring

What problems does our custom solution solve?

  1. Organisation of a complex system of sensor measurements at all hard and soft levels (from sensors to high-level servers), as well as essential organisational and process changes (data management and stewardship, systems for master and measured data management)
  2. Condition monitoring of distributed machinery, critical infrastructure, material supply networks owned or operated by the organisation based on sensor measurements and machine learning techniques
  3. Calculation of material resource balances (supplied - consumed - production needs - losses) within area and time domains
  4. Anomaly detection in infrastructure behaviour based on machine learning techniques and modelling
  5. Generation of alarms for preventive maintenance
  6. Data cleaning, validation, estimation and editing for monitoring needs and data forwarding to complementary services such as billing or workforce management systems

Our approximate custom solution

  1. Organisation of distributed sensor measurement network with automatic data transmission functionality to provide systems with the most relevant and up-to-date information
  2. Analytical infrastructure which includes data integration, preprocessing and storage layers, analytical modules and a data output layer
  3. Data integration layer for batch and real-time streamed data ingestion from the sensor network as well as from other corporate information systems with data cleaning, validation, estimation and editing capabilities
  4. Big data or enterprise data warehouse storage which provides the ability to ingest large volumes of information which is needed to run analytical modules and provide large-scale calculations for monitoring and anomaly detection
  5. Interactive dashboard and reporting system to provide specialists and C-level with the latest and the most important KPIs and detailed information about the state of the infrastructure

Technology

  1. Industrial head end systems and servers for data collection
  2. Industrial master and measured data management systems for data integration, cleaning, validation, estimation and editing
  3. Analytical modules based on machine learning and neural networks for anomaly and loss detection, load forecast and material balance calculation
  4. Interactive web-based dashboard and reporting

Shareholder value impact

Shareholder Value

  1. Revenue Growth
    1. Volume
      1. Acquire New Customers
      2. Retain & Grow Current Customers
      3. Leverage Income Generating Assets
    2. Price Realization
      1. Strengthen Pricing
  2. Operating Margin (after taxes)
    1. Selling General & Administrative (SG&A)
      1. Improve Customer Interaction Efficiency
      2. Improve Corporate/ Shared Service Efficiency
    2. Cost of Goods Sold (COGS)
      1. Improve Development & Production Efficiency
      2. Improve Logistics & Service Provision Efficiency
    3. Income Taxes
      1. Improve Income Tax Efficiency
  3. Asset Efficiency
    1. Property, Plant & Equipment (PP&E)
      1. Improve PP&E Efficiency
    2. Inventory
      1. Improve Inventory Efficiency
    3. Receivables & Payables
      1. Improve Receivalables & Payables Efficiency
  4. Expectations
    1. Company Strenghs
      1. Improve Managerial & Governance Effectiveness
      2. Improve Execution Capabilities
    2. External Factors

FAQ

  • Project pricing & IP related issues

    • What is the approach to projects pricing?

      DAI approach to project pricing is very much flexible and depends on whether DAI has already much experience in the area of client interest or client is ready to collaborate with DAI and investigate new areas of data analytics application. Therefore, DAI has two pricing strategies, traditional and innovative.

      - Traditional model assumes time DAI engineers spend to solve client issues based on existing assets of the institute. Client covers engineer’s time plus the value of the delivered asset customized for the client needs.

      - Innovative business model assumes the co-creation principles that means if client wants to use DAI resources to investigate the area, where DAI and client do not have much experience, DAI is ready to allocate its best resources to solve the client issue. In this case, a significant discount is applied to make the research project attractive for both parties. At the end of such projects, the client has a choice whether to buy out the created IP or share it with DAI. In all cases, pricing is different. Please contact DAI team for more details.

    • What happens with IP?

      DAI pays a lot of attention to the IP rights that appear in projects related to data analytics. In all cases, our clients have the right to buy out the IP that is created during the project or share it with DAI. In all cases, the decision on IP influences directly the project price. Please contact DAI team for more details.

    • Can DAI do project free of charge?

      In case our client has a very good idea and data, DAI can deliver its resources for free to see the idea working with real data and then share the IP for both involved parties. Please contact DAI team for more details.

  • Requirements to Client’s IT Infrastructure

    • Is a quality of client’s data sufficient for project?

      The most part of clients have different kind of problems with data, but the most of these problems are not critical for our solution. We overcome problems with data by application of data mining methods. Clear evidence of sufficient on insufficient of data quality might be estimated during brief data analysis (usually during data journey project). One of result of our projects, we provide recommendations for improvement of data management process for increasing data quality and therefore this leads to increasing solution quality (algorithms prediction, prescription etc.).

    • What are minimal requirements and process for data extraction from client’s IT system?

      At the firs we formulate data model have to be extracted form data. Typically, data is distributed between different IT systems of the enterprise and between different data owners. The first step is to get confirmation for data extraction from IT, security and data owners. The next step is to understand what kind of data is “sensitive” (e.g. personal data, confidential data).

      After all confirmation and limitations are collected, we start data extraction process. We prefer to start data extraction process with client’s forces, if client does not have capabilities and tools we can support him in this process. Sometimes there are significant security limitations on data extraction from client’s systems and it is impossible to get data out of client’s perimeter; in this case, we can work with client’s data by remote connection. There are no limitation systems from which we can get a data, it might be different ERP (SAP, 1C, Oracle etc.), CRM, transactional systems, SCADA system, data warehouses (e.g. Teradata, SAP BW), data lakes (Hadoop) and many others.

    • What are requirements to client’s IT infrastructure for deploying your solution?

      The answer for this question varies for different solutions and client’s IT landscapes. We prefer to deliver solution on premise in client’s virtual infrastructure, but there are possible other variants of solutions. As result, it totally depends on client’s requirements.

  • Solution Development Process

    • What types of project do you deliver (data journey, minimal viable project, and custom solution)?

      There are three types of projects:

      Data journey project. During data journey project, we analyze raw client’s data for business insights extraction and hypothesis exploration. We formulate hypothesis during co-creation with of client’s team experts with Deloitte industry experts and Deloitte Analytics Institute.

      Minimal viable project. During this kind of project, we deliver a pilot working prototype of a product. A prototype shows basic functionality of product and solves most critical client’s tasks. After this project client understands real value of future product.

      Custom solution. This a full development of product included all levels: platform development / backend, algorithms, UI / UX. At the end of this project, a client will get a functionally product solves his tasks.

    • How long does project take time and what are typical project stages?

      Data journey project takes ~2 months. We starts from data audit where we assess data quality. Than we prioritize initial expert’s hypothesizes. After that iterative analyze hypothesis by hypothesis with continues providing results to the client and discussing results. Minimal viable product ~3 months. Development starts with data audit. After data audit we starts development of a product.

      Custom solution from 3 months (in average 7 months). In case of custom solution, we starts from requirements collection, project plan development, next software development and finally quality assurance. The most preferable way is to start from minimal viable product and then develop custom solution.

    • What software development practices do you use for solution delivery?

      During development process, we adhere to the Agile development techniques (Scrum and Kanban). We strongly collaborate with a client in product development. Flowing agile principles of an iterative product delivery, value of the product will be visible on early stages of a development. Client see continues progress on weekly basis. Product development and delivery processes are managed with professional tools for task tracking and project management.

    • Who will participate in a project from client’s side?

      For fully satisfying client’s requirements, we prefer deep client involvement in project, but of course with part-time involvement. There are three main roles from client side. Project customer (project sponsor) who accepts result of project / solution. Business expert who will provide client’s business requirements and supports with client’s specific business expertise. Data / IT expert who will extract data from client side and support team in this field.

  • Delivered Result

    • What results does DAI deliver and what is difference between results of data journey, minimum viable product and custom solution?

      Results of data journey project are insight from data, testing hypothesis and data quality assessment with recommendations for improvement. Result of minimal viable product development is working software prototype, which shows basic functionality for a client (proof of art). Result of custom solution is working software product, which fulfils client’s requirements and operates according to requirements.

    • Do you provide warranty service?

      Yes, we provide warranty service. Exact conditions are individual for each client.

    • Do you teach client’s specialists how to use your software?

      Yes, we teach client’s specialists. Exact conditions are and format of teaching are individual for each client.