Socket

The communication between robot and Vision is based on low level interface on TCP/IP socket. TCP provide stable communication protocol, which is ideal for robot and Vision .

Connections

Type	TCP/IP socket
Port	6969 (TCP)
Byte order	Network order (big endian)

Just like the Client-Server pattern, server is waiting for the requests from client. Vision is the server in this case, so the communication would start from robot. Robot sends a request to Vision , this request should include command and payloads. Vision determines the process by checking command. after processing data on Vision side, it replies back response, this response should include status and payloads. Robot also checks this status, based on this response to preform moving, picking or holding still.

The IP address of server is the IP of workstation which the Vision is running on. Port number must to identical for robot and Vision , otherwise they will not be able to communicate.

Protocols

Request and responses messages between a robot and Vision have a fixed size. A fixed-size message protocol has the advantage that it is easy to implement on the robot side, even with limited programming features.

Note

Even though requests and response are having fix-size length, they do not have starting and/or ending signals. Robot side is able to compare the message size to determine the boundaries of each message.

Requests and response are consist of many fields, each fields are int32 `` size of 4 bytes. Floats would be multiply with a constant number ``MULT = 10000 , then sends back as int32 . After that, client side robot would divide the multifactor to get the data. Negative values would use one’s complement.

Request Message

Content	Type	Length	Description
Position	int32[3]	12 bytes	Robot flange position(x,y,z all in metres) in right hand base. Each position should multiply by MULT factor
Orientation	int32[4]	16 bytes	Robot flange direction in right hand base representation. Each Orientation should multiply by MULT factor
Command	int32	4 bytes	Request command
Payload	int32[2]	8 bytes	Payloads are optional for extra message contents in communication
Meta	int32[2]	8 bytes	meta_1 is the second last field in a message, sending the type of robot to Vision; meta_2 is the last field in message, sending the robot protocol version

Payloads are optional, this field can put anything. However, the other fields are mandatory, without these fields the message is not recognizable by Vision. Command is the main value for vVision to determine which operations to perform.

Response Message

Requests

These are the commands which robot sends to Vision:

RC_DAOAI_NO_COMMAND                             = -1
RC_START_MANUAL_CALIBRATION                     =  1
RC_STOP_MANUAL_CALIBRATION                      =  2
RC_MANUAL_ACCUMULATE_POSE                       =  6
RC_START_AUTO_CALIBRATION                       =  4
RC_AUTO_ACCUMULATE_POSE                         =  7
RC_GUIDANCE_CALIBRATION                         = 10
RC_DAOAI_CAPTURE_AND_PROCESS                    = 20
RC_DAOAI_GET_NEXT_OBJECT                        = 21

RC_START_MANUAL_CALIBRATION

Field	Value	Description
status	DAOAI_MODE_CALIBRATION	Correct Manual Calibration reply, starts the process

RC_STOP_MANUAL_CALIBRATION

Field	Value	Description
status	DAOAI_UNKNOWN_COMMAND	Robot terminates Calibration program

Robot sends the ending signal of calibration mode, Vision would stop calibration mode and sends back DAOAI_UNKNOWN_COMMAND = -1 to acknowledge the termination of calibration mode.

RC_START_AUTO_CALIBRATION

Field	Value	Description
status	DAOAI_MODE_AUTO_CALIBRATION	Acknowledge and enters accumulate mode

Robot sends the start signal of auto calibration mode, Vision would acknowledge the auto calibration mode. Vision would reply DAOAI_MODE_AUTO_CALIBRATION = 11 to acknowledge and enter accumulate mode. If Vision sends back any other status, it means Vision is under different process, robot would resends current command and repeat the above process.

RC_MANUAL_ACCUMULATE_POSE

Field	Value	Description
status	DAOAI_MODE_CALIBRATION	Accumulate mode

Robot requests Vision to accumulate and collect images to accumulates calibration poses. If Vision sends back DAOAI_MODE_CALIBRATION = 10, it means Vision is under correct process. Else, Vision is under a different process and robot will sends RC_START_MANUAL_CALIBRATION = 1 to re-acknowledge the calibration process.

RC_AUTO_ACCUMULATE_POSE

Field	Value	Description
status	DAOAI_MODE_AUTO_CALIBRATION	Accumulate mode
status	DAOAI_DONE_AUTO_CALIBRATION	Collected enough poses, stop Calibration

Robot requests Vision to accumulate and collect images to accumulates calibration poses. If Vision sends back DAOAI_MODE_AUTO_CALIBRATION = 11, it means Vision is under correct process. Else if, Vision sends back DAOAI_DONE_AUTO_CALIBRATION = 33, it means Vision collected enough poses for calibration mode, this status ends the auto calibration mode in robot. If Vision replies anything other than 11 or 33, that means Vision and robot are under different process, robot will resend RC_START_AUTO_CALIBRATION = 4 to restart auto calibration process.

RC_GUIDANCE_CALIBRATION

Field	Value	Description
status	DAOAI_GUIDANCE_CALIBRATION_GOOD	Good pose, proceed to next pose
	DAOAI_GUIDANCE_CALIBRATION_BAD	Bad pose, adjust the pose
	DAOAI_DONE_GUIDANCE_CALIBRATION	Collected enough poses, stop Calibration

Robot requests Vision to start guidance calibration process. Robot would keeps sending this command throughout the guidance process. Vision will start accumulate and collect images to accumulates calibration poses from the first robot pose. Then based on the first pose, Vision would calculate and output the next recommended pose; based on these poses, Vision would reply 12, 13 and 32 according to the calculation: if the pose is good, Vision replies 12, meaning is able to move to next pose; if Vision replies 13, it means the pose is not good according to the calculation; 32 means Vision has collected enough poses to generate calibration result, terminates the calibration mode in robot. If Vision replies anything other than the status above, that means Vision and robot are under different process, robot will resend RC_GUIDANCE_CALIBRATION = 10 to restart guidance calibration process.

RC_DAOAI_CAPTURE_AND_PROCESS

Robot requests the start signal of picking mode, Vision would acknowledge picking mode. Vision would reply DAOAI_DETECTION = 5 to acknowledge and enter detection and picking process. If Vision sends back any other status, it means Vision is under different process, robot would resend current command and repeat the above process.

Response Status

These are the status which robot receive from Vision:

DAOAI_UNKNOWN_COMMAND                           = -1
DAOAI_OBJECTS_FOUND                             =  2
DAOAI_NO_OBJECT_FOUND                           =  3
DAOAI_NO_IMAGE_CAPTURED                         =  4
DAOAI_DETECTION                                 =  5
DAOAI_MODE_CALIBRATION                          = 10
DAOAI_MODE_AUTO_CALIBRATION                     = 11
DAOAI_GUIDANCE_CALIBRATION_GOOD                 = 12
DAOAI_GUIDANCE_CALIBRATION_BAD                  = 13
DAOAI_DONE_GUIDANCE_CALIBRATION                 = 32
DAOAI_DONE_AUTO_CALIBRATION                     = 33

Message metadata

Fields	Description
Meta_1	ABB = 0
	Fanuc = 1
	Hanwha = 2
	Hanwha = 2
	Kuka = 3
	Omron_TM = 4
	Omron_TM = 4
	Siemens_PLC = 5
	Staubli = 6
	Staubli = 6
	UR = 7
	Yaskawa = 8
	Efort = 9
	Efort = 9
	Aubo = 10
	Dobot = 11
	Dobot = 11
	Mitsubishi = 12
	Other = 99
Meta_2	This is the protocol between Vision version and robot version, in this case meta_2 is 1.

If your robot does not follow the protocols above, Vision would report errors, and terminates any process in order to protect the robot.