AI Rd-03E Manuel utilisateur

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 1 of
18
Rd-03E Gesture Recognition
User Manual
Version
V1.0.0
Copyright
©2023

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 2 of
18
Document resume
Version
Date
Develop/revise content
Edition
Approve
V1.0.0
2023.9.14
First formulation
Zhanyun Lin
Zifeng Chen

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 3 of
18
Introduction
This document describes the basic functions, hardware specifications, software
configuration, and installation conditions of the Rd-03E gesture recognition control reference
design XenG102ST. It aims to help developers quickly get started with human sensing and
gesture recognition solutions based on Xen102 hardware, easily configure parameters that are
most suitable for their own application scenarios, and build their own accurate gesture
recognition control sensors.

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 4 of
18
Content
1 Rd-03E Overview...................................................................................................................... 5
2 System description .................................................................................................................... 5
3 Hardware Description ............................................................................................................... 6
4 Software Description................................................................................................................. 7
5 Communication protocol........................................................................................................... 9
6 Installation instructions ........................................................................................................... 10
7 Detection range ....................................................................................................................... 11
8 Rendering ................................................................................................................................ 13
9 Gesture Description................................................................................................................. 13
9.1 Gesture Definition........................................................................................................ 13
9.2 gesture speed ............................................................................................................. 14
9.3 gesture parameters........................................................................................................ 14
10 Considerations....................................................................................................................... 15
Contact us................................................................................................................................... 16
Disclaimer and copyright notice ................................................................................................ 17
Notice ......................................................................................................................................... 17
Statement.................................................................................................................................... 18

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 5 of
18
1 Rd-03E Overview
Rd-03E gesture recognition control reference XenG102ST design with extremely
simplified 24 GHz radar sensor hardware Xen102 and smart algorithm firmware ST01.
The Xen102 hardware consists of an AIoT millimeter-wave radar chip S3KM111L, a
high-performance one-transmitter-one-receiver microstrip antenna, a low-cost MCU and
peripheral auxiliary circuits. Smart Algorithm Firmware ST uses FMCW waveform and
S3KM111L chip proprietary advanced signal processing technology. This reference scheme is
mainly used in intelligent bathroom scenes, which can realize accurate human body sensing
ranging and gesture recognition. Developers can display the distance information and gesture
information in real time through the host computer tool.
The radar beam of Xen102 hardware adopts wide and narrow design, and the angle range
of antenna 4 patch direction is ± 20 °, and the angle range of vertical antenna 4 patch direction
is ± 40 °. The user can adjust the antenna placement direction during installation according to
the needs of the scene. In toilet applications, it is recommended to use a horizontal narrow
beam to avoid false triggering by passing people around.
2 System description
Rd-03E is a gesture recognition control reference design based on silicon
micro-S3KM111L chips. The sensor uses FMCW waveforms, combined with the MCU's
proprietary radar signal processing and built-in intelligent presence sensing algorithms to
detect targets and gestures in a specified area and report results in real time. Based on this
reference solution, developers can quickly develop corresponding induction products. Table
2-1 shows the specification parameters of the Rd-03E.
Table 2-1 Rd-03E Specifications
Parameters
Minimum
Typical
Max.
Unit
Working frequency band
24.015
-
24.245
GHz
Sweep bandwidth
-
0.23
-
GHz
Maximum equivalent
omnidirectional radiated
power
-
12
-
dBm
Supply voltage
4.6
5
5.4
V
Dimensions
-
28x24
-
mm2
Ambient temperature
-40
-
85
℃

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 6 of
18
Average operating current
-
50
-
mA
Detection distance
0.3
-
2.2
m
Detection angle (H plane)
-
±40
-
°
Detection angle (E plane)
-
±20
-
°
ranging error
-
-
±10
cm
3 Hardware Description
Rd-03E supported hardware is Xen102. Figure 3-1 is a photo of the front and back of the
Rd-03E. Five pin holes J2 are reserved for the Rd-03E hardware (no pin is provided at the
factory), which are the power supply and communication interface. The MCU burning port is
called J3. Please connect according to the corresponding pin name when burning. (J1 is
reserved for USB interface, which is not introduced in this document for the time being.)
(a)front (b)back
Figure 3-1 Rd-03E hardware physical diagram
The pin descriptions of J2 and J3 are shown in Table 3-1 and Table 3-2.
Table 3-1 J2 Pin Description
J#PIN#
Name
Function
Operating Range
J2Pin1
VCC
Power input
4.5V~5.5V,Type5V
J2Pin2
GND
Ground
-
J2Pin3
OT1
UART_TX
0~3.3V
J2Pin4
RX
UART_RX
0~3.3V
J2Pin5
OT2
UART_TX(reserved)
0~3.3V
Table 3-2 J3 Pin Description
J#PIN#
Name
Function
Operating Range
J3Pin1
GND
Ground
-
J3Pin2
DIO
Data port
0~3.3V
J3Pin3
CLK
Clock signal
0~3.3V

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 7 of
18
J3Pin4
3V3
Power input
3.3V
Rd-03E supports keil 5 IDE to burn hex files or source code projects. You can use J-Link
(V9 or above), CMSIS-DAP and other burners to download programs. Please make sure it is
installed before burning. GiqaDevice.GD32E23x_DFP.1.0.1.pack and ARM.CMSIS.
5.7.0.pack or a later version of the CMSIS pack.
4 Software Description
Rd-03E the hardware uses 5V power supply. When connecting hardware and serial port
tools, attention should be paid to the connection mode of hardware and serial port tools: TX of
hardware is connected to RX of serial port tools, RX of hardware is connected to TX of serial
port tools, and jumpers are connected to 5v power supply, as shown in fig. 4-1. If you use other
serial port tools, also need to meet the 5V power supply.
Figure 4-1 Schematic diagram of Rd-03E module and serial port connection
After connecting the hardware, serial port tool and upper computer as shown in Figure 4-1,
open the upper computer software ICLM_XenG102STTool and configure the module
according to the following steps:
Step 1: Open the upper computer software, click the "Refresh Serial Port" button in the
upper left corner of the interface (Figure 4-2 marked area 1), if the distance curve graph
appears below the interface, it indicates that the connection is successful;
Step 2: After the module is successfully connected to the upper computer, select the
corresponding serial port number in the upper left corner of the interface (No. 2 area marked in
Figure 4-2) and enter the baud rate, fill in the reporting cycle according to the reporting
frequency of the firmware, and the reporting cycle of the factory firmware is 50 ms;
Step 3: Click the "Start/Stop" switch button (area 3 marked in Figure 4-2) to receive
real-time data from the radar and view the real-time waveform of the target distance at the
bottom of the upper computer interface. The window above the interface (area 4 marked in
Figure 4-2) can display the distance information of the target and the toilet status in real time.

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 8 of
18
Figure 4-2 Upper computer tool ICLM_XenG102STTool configuration interface
The distance information can divide the gears by setting different distances. The upper
computer tool divides the sensing distance into 3 gears, which are respectively represented by
colored lights of different positions and colors. The display principle is as follows:
⚫When the target distance is between 0cm and 60cm, the leftmost red light is on, and the
lights in other gears are off;
⚫The yellow light in the middle of the target distance of 60cm and all 130cm is on, and
the lights on both sides are off;
⚫The green light on the right side of the target distance of 130cm to 200cm is on, and the
two lights on the left are off.
The target distance interval and the on-off state of the colored lights are shown in Figure 4-3.
The distance interval corresponding to each gear can be customized through the upper computer
tool interface.
Target distance interval(cm)
Light zone display
Zero~ Sixty
Sixty~ One Hundred Thirty
One Hundred Thirty~Two Hundred
Figure 4-3
Toilet cover ring 5 different states, respectively:
⚫No one (the cover ring is closed by default);
⚫Open the cover by a person (when the cover is closed);

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 9 of
18
⚫Gesture flip cover (in open state);
⚫The cover is closed by people (when the cover is opened);
⚫Close the cover when the person goes through the closing circle (in the state of opening the
cover circle);
A schematic diagram of the above state is shown in FIG. 4-4.
(a) No one state (b) People come to open the cover (c) Gesture circle turning
(d) people go to close the cover (e) people go to close the cover
Figure 4-4 Schematic Diagram of Toilet Status
The unmanned delay refers to the time interval for delaying the report of the unmanned
result when the radar self-detects the unmanned state. Developers can adjust the parameter by
modifying the parameter NobodyDebounceTime = "1000 ° in the
XenG102STTool/appConfig.xml file under the root directory of the host computer tool, and its
default value is 1000 ms.
5 Communication protocol
This communication protocol is mainly used by developers who need to separate from
visualization tools for secondary development. The Rd-03E hardware communicates with the
outside world through the serial port. The serial port outputs the radar data processed by the
algorithm by frame. The format of a frame of data is:
AA + distance information (two bytes in cm) + gesture information (1 byte) +55
The distance information is in little-end format, and the data example is shown in Figure 5-1.
The output delay of the gesture result is about 0.25s.
Radar serial port data parsing distance is as follows:

Rd-03E Gesture Recognition User Manual V1.0.0
Copyright © 2023 Shenzhen Ai-Thinker Technology Co., Ltd All Rights Reserved
Page 10 of
18
AA 2D 00 00 55, then distance = 002D = 45cm, gesture = 00, no gesture.
AA 36 00 01 55, then distance = 0036=54cm, gesture = 01, gesture.
The distance information can be displayed by the host computer tool, as shown in Figure
5-1. The top of the software interface shows the target distance measured by the millimeter
wave radar, the far right is the toilet state, and the red curve below the interface is the real-time
distance waveform.
Figure 5-1 Rd-03E gesture recognition PC tool interface diagram
6 Installation instructions
When installing the module on the bare board, the installation position should be 45cm
higher than the ground, and the placement angle should be 45 upward, as shown in Figure 6-1.
When installing, pay attention to the antenna direction, and the antenna 4 patch direction
should be taken as the horizontal direction.
When the module is installed on the toilet, the recommended installation angle is the same
as that of the bare board installation, which is 45 ° upward.
Table des matières

















