Monday 24 January 2022

What is AI (artificial intelligence) required for project management tools?

Business (BtoB) has been using AI (artificial intelligence) more and more. Since Google announced ai (artificial intelligence) that recognizes cats in 2012, it has finally become available in business (BtoB) for about 7 years. There are two main types of AI that are practical at the business level: "AI that you want to make a thing" and "AI that can be made into things".

1) Natural language understanding AI 2) Image recognition AI And there are also speech recognition and text analysis, but the number of cases of these two AI functions is increasing at an incredible rate. Let's take a look at what practical examples there are.

AI (artificial intelligence) "AI of natural language understanding function" and "AI of image recognition function" used in business (BtoB)


it is a function that learns various data and gives answers to guide you in the right direction. in other words, the function that performs "prediction" that supports human judgment is good in understanding natural language.

< Natural Language Understanding Function Use case >

・ AI service

that derives the evaluation value faster than a real estate appraiser This AI function instantly analyzes the size and market price of land and buildings, market analysis, and regional analysis to calculate the value of real estate. Ultimately, the real estate appraiser determines the valuation value, but if it is performed by a real estate appraiser, it will be calculated in 1 hour and AI in 5 minutes. I feel that the combination of natural language comprehension functions that combine rule-based machine learning and black box deep learning is very suitable.

In addition, there are AI assistants and natural language understanding functions that provide secretarial support, but I think that level is not yet something that can be used in business (BtoB). If you liken it to the student level, it can be said that it is an elementary school student or less. Predicting AI is a very wanted function, and there is no doubt that it is an AI that you want to make a thing. However, ai that predicts the current level of natural language understanding function and the problem of developing learning data is still difficult, and it seems that it is a long way off to become a thing in business (BtoB).

2) ai (artificial intelligence) of image recognition function image recognition is an ai function that can improve the quality by deriving correct images by learning various images. at present, it is more used in business scenes than natural language comprehension functions.

< Image Recognition Function Use Case study >

  • Used car assessment AI service Not only learning of the year and model of used cars, but also learning the scratches and dents of the appearance by image recognition greatly shortens the time spent on used car assessment performed by humans, leading to quality improvement.
  • Manufacturing: Product appearance inspection and anomaly detection AI service By learning good and defective products of products with AI rather than rule-based visual inspection and abnormality detection systems, image recognition functions greatly improve quality and reduce visual labor
  • Costs.

In addition, ai (artificial intelligence) with image recognition functions, such as inspection of the appearance of construction sites and security functions by facial image recognition, is overwhelmingly more common in business (BtoB). In other words, it can be said that it is "AI that can be made into things".


We found that there are many cases of use in business (B2B) using image recognition, but what kind of AI is required in the field of project management tools?

1) When

using natural language understanding function AI (artificial intelligence), "prediction" can be done, so it is often said that "I want you to predict projects that will be in the red if you continue as it is", "I want you to calculate the success rate at the planning stage", etc., "I want you to predict and predict and give advice about future projects!" However, the natural language understanding function is still at the elementary school level, and there is no learning data in the company that can make such predictions in the first place.

If there is past data of "company" of successful or failed projects, and past data of "other companies" collected in the cloud, it may still be feasible, but I think we can still go to such a stage. However, there is no doubt that it is "AI function I want if there is". I want it, too, and with this feature, project management tools: OBPM won't sell explosively!" I strongly think. (Laughs)

using image recognition function AI (artificial intelligence), the strength of image recognition is to change "unstructured data" to "structured data". Unstructured data refers to images of products, human faces, handwritten characters and pictures, etc. The key point of image recognition function is that it can be used in business by turning them into data with the power of AI. So what do you want to turn into structured data in project management: what is unstructured data in project management?


What kind of scenes are coming out of project management where you want to turn unstructured data into structured data with image recognition?

1) Man-hours, cost control, and

quality control data managed in Excel Since the format is managed separately by each individual and team, you can understand if you want uniform data. However, these are already "structured data" and there is no need to use image recognition AI functions.

2) Gantt chart mentioned in Excel

It is structured data because it is also Excel. However, compared to man-hours, cost control, and quality control data, there may be a need to make structured data with image recognition AI only there, because there is an image of a chart called a Gantt chart as well as numerical data.

Currently, structured Gantt charting tools are the most popular for Microsoft MS projects, and if they are in the file format of MS projects, data can be read into Gantt charts. However, it is difficult to make Excel's Gantt chart structured data, so it may be interesting to use the image recognition AI function.

3) If the system has already been developed and you want to create a system design document

later There are many scenes in project management that perform deliverable management. The most time-consuming deliverables are those related to the system screen, such as "screen layout definition" of basic design books and detailed design books. For example, for an aging legacy system that has been in use for many years, you may want to prepare a system design document in order for an expert to retire, or you may want to prepare a basic design document or detailed design document later because you have advanced system development in agile development. By making such a system screen an image, it can be read with the image recognition AI function. It also fits the concept of image recognition AI, "First of all, there is unstructured data, and turning it into structured data is the strength of AI".

ai (artificial intelligence) required by project management tools is now an image recognition function image recognition ai described in the above 3) has already been sold as a "tool to support system design document with image recognition ai function", and has been introduced by many companies.

"ObjectBrowser Designer",

a CAD tool that supports system development design  documentation,

However, this is one of the "system deliverable management" in the project management tool, and it is not yet possible to cover the entire project management with AI (artificial intelligence). As the No. 1 company, we aim to introduce AI functions that predict and predict from project planning to results, such as functions that turn unstructured data into structured data necessary for project management tools and natural language understanding AI functions using image recognition AI functions.

Currently, I would like to realize AI functions that can be used in the business (BtoB) world and are required for project management, rather than "Nancha-na na AI functions: project management tools" released by other companies.

No comments:

Post a Comment