Easily develop programs for learning IA robots using Ruby, Java, PHP, and C#.
IA robots work just like real humans, using PC screen recognition, camera object recognition, mouse & keyboard input.
Realization of an autonomous distributed grid computing network that enables two-way communication between IA robots.
At our company, we call software robots equipped with AI that can analyze situations, make judgments, and automatically perform prescribed tasks in the same way as humans do IA(Intelligent Automation) robots.
An IA robot is a robot that uses low-code to forcibly learn tasks that humans have previously performed in their work, and then performs the tasks on behalf of humans.
The learning method of our AI robot is classified as forced learning, a type of supervised learning in AI terms.
Our IA robots are robots that use low code to forcibly learn tasks that humans already have work experience performing on behalf of humans.
In AI, the information obtained by machine learning and deep learning is prediction or inference and is not accurate.
Using techniques such as machine learning to train robots to perform specific tasks can be risky, and training a robot using machine learning requires large amounts of data and time.
Business robots are required to accurately and properly carry out the work of specified commands.
Forced learning, a type of supervised learning, is a fast and accurate way to teach human know-how directly to a robot.
Our IA robot concept is that IA robots work autonomously (Standalone) and automatically by learning the decisions analyzed by humans (skilled and expert workers).
The software that can be developed to transform a PC into an IA robot is the
Robowiser Framework (RBF).
The concept of the IA robot is to automate IT tasks by making the robot learn human operations and decisions as they are, easily and directly by anyone.
RBF began with the development of Runbook Automatuon (RBA) software robots around 2007, expanded its functions from Robotic Process Automation (RPA) to AI, and now has become a high-level software robot that incorporates the three elements (DNA) of RBA, RPA and AI.
RBF provides a robotic API that can flexibly handle a variety of complex human decision-making and IT tasks.
For example, it can instantly select multiple options by combining a large amount of information obtained by recognizing PC resources and screen displays, automatically operate the mouse and keyboard, and easily develop the same complex decision-making operations as humans in low code.
Our Robowiser Framework products are designed for environments where no one is in front of the PC, and are being developed based on the concept of realizing unmanned operation. We conceive, design, develop, and provide various APIs to control software robots in order to realize unmanned operation.
The actual five senses are sight (seeing), hearing (listening), taste (tasting), smell (smelling), and touch (feeling with skin), RBF can replace these with ① thinking, ② seeing, ③ speaking, ④ hearing, and ⑤ manual input, as shown in the figure below, to develop the senses replaced by RBF.
Communication in human work is achieved through two-way communication (conversation) from person to person.
Similarly, two-way communication from robot to robot is necessary for all people in an organization to cooperate and get the job done.
Therefore, we thought that a communication network for multiple people to work is almost the same as a grid network.
A grid network is a type of distributed network in which multiple computers and devices are interconnected and work together to complete tasks.
For this reason, RBF has adopted and implemented the concept of grid networking in its design, which allows IA robots to be placed on a grid network.
This allows you to replace employees with IA robots in the same functional locations within your organization.
・Two-way communication between IA robots
・Communication relay between IA robots
・NAT traversal in communication between IA robots
・Reverse NAT in communication between IA robots
We have achieved this.
Communication data is encrypted, data is compressed, and connections between IA robots are authenticated.
Grid network communication can be easily developed by simply specifying RBF API arguments.
The advantages of placing IA robots on this grid network are as follows:
1. Scalability
The grid network improves the scalability of the entire system because multiple IA robots can cooperate and share tasks.
Processing capacity can be easily expanded by adding new IA robots.
2. Efficient resource utilization
Each IA robot can share specific tasks, optimizing resources.
This reduces energy consumption and time waste.
3. Redundancy and reliability
The grid network allows multiple IA robots to perform the same task, so even if one IA robot fails, the entire system can continue to function, improving reliability.
4. Real-time data processing
Multiple IA robots can collect and process data simultaneously, enabling real-time decision-making.
This allows them to function effectively in situations where a rapid response is required.
5. Collaborative work
By sharing and cooperating with each other, IA robots can perform more complex tasks more efficiently and achieve results that could not be achieved by a single IA robot.
6. Flexibility
Grid networks can integrate different types of IA robots with AI systems, allowing for flexible operation for a variety of purposes.
IA robots are equipped with edge AI camera functionality.
Edge AI camera indicates that data processing for object recognition is not performed by the cloud, but by the IA robot at the end of the line.
By machine learning the data annotated with objects requested by customers, it can recognize, detect, and detect objects such as people, cars, and faces.
It can also inspect the appearance of objects by forcing it to learn the object's features.
Applications are expected to be broadened by deploying multiple [IA robots + edge AI cameras] on a grid network.
You can easily control IA Robots by simply setting API arguments.
- | Keyboard input / Mouse operation / Mouse drag and drop |
- | There are five ways to get screen display information. 1:Window text capture 2:Object text characters in window memory 3:Graphic patterns 4:OCR 5:Screen coordinates |
- | Windows screen control * Log on / Log off * Screen lock / unlock |
- | Receive mail lists in your mailbox, send and receive mail, and send and receive attachments from your mail server. |
- | Text characters are read out loud (voice) over a speaker and controlled by human voice input to a voice recognition engine. |
- | Obtaining location information from location-based services (e.g., GPS sensors) |
- | Controls power supply for PCs and monitors. |
- | File transfer between IA Robots. |
- | Cross the IP segment, Reverse NAT, Communication relay between IA Robots. |
- | Cooperation with IoT devices. |
- | Creating and Editing Excel files. |
- | Edge AI cameras Appearance Inspection or detect people, vehicles, objects, and more. |
Software that supports integrated system operation by IA Robots and development of IA Robots
Management software that can manage programs that control IA Robots by registering programs developed with RBFAPI as job commands.
Job commands can be executed on a specified schedule.
The RBF Application Analyzer is used to develop RBF APIs.
It is software that can analyze the operation of the screen & application to be controlled in advance and simulate the automatic control of the keyboard and mouse.
RBF QuickRobo records human operation and run recorded operation. The recorded operation can be run using the RBF API.
This software can be expected to reduce the number of days to development.
- | Monitoring the operating status of multiple IA Robots. |
- | Monitor the status of the Windows screen of the IA Robot(PC) on the grid network. The status of the Windows screen is to monitor * Windows logon screen * Lock screen * Desktop screen * Remote desktop screen changes. When the screen display changes, AI Agent Robot Controller can notify you by email. |
- | AI Agent Robot Controller can run programs file or executable file in sequence on multiple IA Robots. Program files or executable files can be transferred to the IA Robot for run. Run a program file or execute file that exists in the IA Robot. |
- | Notify you of the execution result of the program file or execute file by e-mail. |
- | Graphical display of trends in program execution results. |