Linoc: A Prototyping Platform for Capacitive and Passive Electrical

Field Sensing

Julian Von Wilmsdorff

, Malte Lenhart

, Florian Kirchbuchner

and Arjan Kuijper

Fraunhofer Institute for Computer Graphics Research IGD, Darmstadt, Germany

Technische Universit

at Darmstadt, Darmstadt, Germany

Keywords:

Capacitive Sensing, Sensor Toolkit, Rapid Prototyping, Development Board, Electric Field Sensing, Electric

Potential Sensing, Embedded Programming, Sensor.

Abstract:

In this paper the Linoc prototyping toolkit is presented. It is a sensor toolkit that focuses on fast prototyping

of sensor systems, especially on capacitive ones. The toolkit is built around two capacitive and two Electric

Potential Sensing (EPS) groups providing unobtrusive proximity detection in the ﬁeld of Human Computer

Interface (HCI). The toolkits focus lies on its usability and connectivity in order to be adapted in future research

and novel use cases. A common obstacle in the beginning of a project is the time required to familiarize with

present tools and systems, before the actual project can be attended to. Another obstacle while tackling new

tasks is the actual physical connection of sensors to the processing unit. This situation can be even worse

due to dependencies on previous work, most of the times not fully documented and missing knowledge even

if the the original designer is involved. Good toolkits can help to overcome this problem by providing a

layer of abstraction and allowing to work on a higher level. If the toolkit however requires too much time to

familiarize or behaves too restrictive, its goal has been missed and no beneﬁts are generated. To assess the

quality of the Linoc prototyping toolkit, it was evaluated in terms of three different aspects: demonstration,

usage and technical performance. The usage study found good reception, a fast learning curve and an interest

to use the toolkit in the future. Technical benchmarks for the capacitive sensors show a detectable range equal

to its predecessors and several operational prototypes prove that the toolkit can actually be used in projects.

1 INTRODUCTION

When it comes to prototyping with any kind of sen-

sors, the most common ones can be connected via

intra-board bus protocols such as Inter-Integrated Cir-

cuit (I

C), Serial Peripheral Interface (SPI), Univer-

sal Asynchronous Receiver-Transmitter (UART) or

JTAG-programmers. This often requires program-

ming skills and knowledge of embedded systems in

order to work with the sensor data. Many sensor sup-

pliers attempt to reduce the amount of programming

needed, by supplying libraries or other interfaces to

open source platforms like Arduino or Raspberry Pi.

However these platforms or libraries often lack

advanced features and prevent programmers to access

many in-depth parameters or functions, making them

unsuited for more complex projects.

Rapid prototyping has made its way into research

and design processes and accelerates the adoption of

new techniques and concepts. The Linoc rapid pro-

totyping toolkit is designed to provide an easy to use

platform to be used in future projects for easy data ac-

quisition. With this in mind the requirements for the

ﬁrmware of the Linoc prototyping toolkit are on one

hand the ease of usage and on the other hand the pos-

sibility for further reﬁnement for advanced use cases.

The board has two measurement groups each for ca-

pacitive sensing and for passive EPS. Both measure-

ment principles are primarily used to detect activity,

proximity and movement, each method performing

differently depending on the ambient conditions. It

is assumed that the Linoc board in most cases will be

used mainly as a sensor whereas signal processing as

well as higher level interactions are done on a separate

computer. For this the setup and conﬁguration proce-

dure needs to stay as simple as possible. If then for

example a demonstration was successful and the next

step is to eliminate the need for a separate computer,

the ﬁrmware design needs to allow modular exten-

sions to be integrated. Only at this point knowledge

of C programming and embedded systems is required.

The aim of the Linoc prototyping toolkit is to provide

easy interfaces for data collection, setup of sensor net-

works and conﬁguration of the board, while preserv-

Von Wilmsdorff, J., Lenhart, M., Kirchbuchner, F. and Kuijper, A.

Linoc: A Prototyping Platform for Capacitive and Passive Electrical Field Sensing.

DOI: 10.5220/0010336400490058

In Proceedings of the 10th International Conference on Sensor Networks (SENSORNETS 2021), pages 49-58

ISBN: 978-989-758-489-3

Figure 1: Front and back of the Linoc prototyping toolkit.

ing the possibility to easily customize the source code

to satisfy advanced use cases. To obtain larger sen-

sor arrays, multiple Linoc boards can be connected to

form a sensor network. One device will then take the

role of the master to aggregate sensor data of multi-

ple slaves before posting them to the server or host

computer via USB or a wireless technique. This pa-

per begins with an overview of the role of toolkits

in the modern design process, presents related work

in the ﬁeld of HCI and proximity sensing alongside

the actual hardware design of the Linoc toolkit. The

Linoc hardware is detailed in section 4 followed by

the ﬁrmware design and its components, alongside

information about challenges, limitations and design

choices during the implementation process. In the ﬁ-

nal chapter the toolkit is evaluated and an example

project is presented, demonstrating the toolkit’s fast

prototyping capabilities. This is accompanied with a

use case study to determine its usability.

2 RELATED WORK

There are a couple of toolkits in the HCI area avail-

able. In this section, we will summarize them

and compare them with the Linoc rapid prototyping

toolkit.

Hamblen et al. (Hamblen and van Bekkum, ) de-

scribe the process of building a prototyping toolkit

for future student works. They provide a cloud based

compiler to their students to eliminate setup time and

provide documentation in form of a wiki. The cloud

based compiler is an attractive way to eliminate plat-

form dependencies and can be hosted cheaply on even

weaker computing platforms like single board com-

puters.

WatchConnect (Houben and Marquardt, ) is a

toolkit to develop cross-platform applications to ex-

plore interactions between the smartwatch and a sec-

ond screen. The sensor data as well as the smart-

watch’s screen are used to extend classic input meth-

ods.

The Proximity Toolkit (Marquardt et al., ) pro-

vides an open-source hardware setup, interfaces to ac-

cess higher-level proximity representation and a tool

for visualization to use in proximity aware applica-

tions. It is designed to be hardware oblivious, so that

different sensor techniques can be used. It focuses

on the interaction with digital devices and provides

developers with information about “orientation, dis-

tance, motion, identity and location information be-

tween entities”, which are the dimensions for proxim-

ity in ubiquitous computing deﬁned by Greenberg et

al. The authors state that their motivation is to address

the initial problem to acquire proximity sensor data

for developers. Even if sensing hardware is available,

effort is still required to translate sensor information

into proximity, as calibration and noise can have a big

impact. Their setup differs from the Linoc prototyp-

ing toolkit context as this toolkit is designed to require

no counterpart in sensors or devices to work. Further-

more Linoc aims to be unobtrusive whereas the prox-

imity toolkit focuses on intended interactions.

Midas (Savage et al., ) is a toolkit to design ﬂexi-

ble capacitive touch sensors to apply to other objects

in order to enrich interaction possibilities. While their

focus lies on touch interactions it might be interesting

to use their toolkit for electrode design. The evalua-

tion was done in a way that the participants received a

task to complete. They received feedback that videos

can convey certain instructions better than images and

addressed this by adding animations.

The CapToolkit (Wimmer et al., ) is the second

generation of capacitive toolkits by Wimmer et

al. A stated goal is to make implicit interaction

concepts easier to develop. It supports up to eight

loading mode sensors, as shown in Figure 2 with

a sampling frequency between 25Hz and 100Hz.

SENSORNETS 2021 - 10th International Conference on Sensor Networks

Figure 2: Captoolkit by Wimmer et al (Wimmer et al., ).

Sensor reconﬁguration is possible at run-time using a

custom protocol via USB connection. User Datagram

Protocol (UDP) and Transmission Control Protocol

(TCP) interfaces can be used by the connected host

computer. The Linoc prototyping toolkit provides

this functionality directly by the sensor board, which

is made possible by recent microcontroller develop-

ment. Thus these interfaces can still be used with

battery powered sensors without physical connection

to a computer. With a 10cm x 10cm electrode the

CapToolkit is able to detect a human body at a dis-

tance up to 1m and hand movements up to 50cm. The

spatial resolution is given at 1cm at 25cm distance,

but was not reproduced in experiments by Puppedahl

et al. (Grosse-Puppendahl et al., a).

Figure 3: OpenCapSense by Puppedahl et al. (Grosse-

Puppendahl et al., b).

The OpenCapSense (shown in Figure 3) board is

inspired by the CapToolkit and addresses three short-

comings: conﬁnement to loading mode capacitive

measurement, slow sample frequency and no options

to connect multiple sensor boards. The last aspect

is overcome by providing two Controller Area Net-

work (CAN bus) real time bus interfaces to synchro-

nize data. The board features eight Universal Serial

Bus (USB) ports for external capacitive sensors and

a framework for data exchange. It supports different

sensor types, and has been used with the three differ-

ent capacitive measuring modes, as well as EPS sen-

sors (Grosse-Puppendahl, ). It operates at frequencies

up to 250Hz with eight and up to 1kHz with one sen-

sor attached. (Grosse-Puppendahl et al., a).

The OpenCapSense board is in a way the direct

predecessor of the Linoc prototyping toolkit and has

been used in various research projects (Braun, ) in

the HCI area. It was originally designed by Tobias

Große-Puppendahl. It is thus closest related in appli-

cation purpose, albeit having a different sensor con-

cept. Linoc also focuses even more on usability than

OpenCapSense and was designed with a more modern

processor and more possibilities to connect the sensor

to a computer.

3 DESIGN GOALS

Before describing the actual hardware and software,

we will brieﬂy discuss the design goals of the Linoc

prototyping toolkit. Linoc was built to be ﬁtting for a

wide variety of use cases. That is why it should meet

the following requirements.

1. High connectivity

2. Easy to use

3. Support for multiple programming languages

4. Realization of advanced use cases possible

5. Chaining of multiple boards

Design goal 1 was the main reason to create Linoc.

While many available toolkits have multiple inter-

faces, most of the time additional hardware is needed

if more than a USB connection is required. This leads

to more effort required while building an actual pro-

totype both for hardware and software development.

An easy to use user experience as stated in design

goal 2 is very important if the toolkit is to be used in

an academic environment. This will enable students

and teachers alike to use Linoc seamlessly in every

kind of project. It also reduces the amount of time

that users need, both experienced and inexperienced

ones, to get their hands on the ﬁrst data.

While most toolkits require a certain program-

ming language to use them, Linoc was designed so

that it can be programmed in several languages. In

times where new programming languages are created

apparently on a daily basis, it is another time consum-

ing hassle to be bound to a certain language that the

user might not prefer. With the freedom to choose the

language as stated in goal 3, the user is empowered

to use the language he is most capable of and that is

most ﬁtting for his current project.

Having a toolkit that is easy to use should not limit

the number of use cases you can use it for. In many

cases, easy usability restricts versatility. To prevent

Linoc: A Prototyping Platform for Capacitive and Passive Electrical Field Sensing

Linoc from being a prototyping toolkit only for be-

ginners, we included design goal 4.

Another important aspect learned while using

other toolkits was, that no matter how many sensors

can be attached to a board, there is always a use case

where you need more. While some toolkits like Open-

CapSense have a CAN-bus interface to combine mul-

tiple boards, it is still a task that takes up several hours

to connect and reprogram them. That is why we added

design goal 5. In this way, the user is able to acquire

higher dimensional data by adding more sensors.

After explaining why we choose these design

goals, we will now focus on how we implemented

them in hardware and software.

4 HARDWARE

4.1 Microcontroller

The microcontroller used on the Linoc board has to

fulﬁl several requirements; The choice of the micron-

controller is a critical step because most of these re-

quirements have to be covered by the controller itself.

To be precise, the controller affects the design goals 1,

3 and 4. The ESP32 from Espressif fulﬁlls all of these

requirements. It features Wi-Fi, Bluetooth and Eth-

ernet connectivity as high level protocols. For intra-

board communication, the user can choose between

several low level protocols such as UART, I

C, Inter-

IC Sound (I

C), CAN bus and SPI to name only the

most common ones. This makes the ESP32 a good

processor for our needs of high connectivity. The

ESP32 has enough memory to support a Real-Time

Operating System (RTOS). This allows for complex

use cases as we planed in goal 4.

Since the ESP32 is the direct successor of the

ESP8266, which was (and still is) very popular in the

maker-scene and for Internet of Things (IoT) prod-

ucts, there are a couple of toolchains for different pro-

gramming languages available.

Table 1: Programming languages and toolchains available

for the ESP32.

Toolchain Language Editor

ESP-IDF

C/C++

with FreeRTOS

Eclipse IDE

Mongoose OS C & JavaScript Browser page

MicroPython Python -

PlatformIO C/C++

Atom editor,

Eclipse IDE,

Visual Studio Code,

...

Arduino C++ Arduino IDE

Table 1 lists a selection of possible languages and

editors for the ESP32. Note that this not only is in

favor of our design goal 3, but also goal 2. This is

because the ESP32 is, amongst others, also Arduino

compatible. Arduino is a commonly used language by

beginners when it comes to embedded programming.

4.2 Capacitive Measurements

The Linoc prototyping toolkit features two capacitive

loading mode sensors. The sensors were added di-

rectly onto the board, so that in contrast to other pro-

totyping toolkits, there is no need to add more exter-

nal hardware.

Figure 4: Model of the capacitive sensor.

The capacitive group’s main component is the 555

timer which generates the charging cycles on the elec-

trode, as shown in Figure 4, with a constant charg-

ing current. Since the current is constant, the charg-

ing process is inﬂuenced by the environment. When

a conductive objects approaches the electrode, the

resulting capacity between electrode and object in-

creases. Hence, the time needed to charge this ca-

pacitor with the same amount of current increases.

A pulse counter module in the microcontroller then

counts the number of cycles per second, which is the

ﬁnal measurement parameter. The microcontroller is

also able to dis- and re-enables the 555 timers so that

it is possible to prevent cross talk when multiple sen-

sors are active at the same time.

Linoc also features a shield for both capacitive

sensors to support coaxial cable or to minimize en-

vironmental inﬂuences on selected parts of the elec-

trode. The shield is realized by a simple voltage fol-

lower, implemented with an opamp, of the capacitive

feed line.

4.3 Electric Potential Sensing

Besides the two capacitive sensors, Linoc features

two electric potential sensors. Electric potential sens-

ing is another form of capacitive sensing. The main

differences between a capacitive loading mode sen-

sor and an electric potential sensor are the detection

range, the mode of operation and the energy con-

sumption. Electric potential sensing can detect mov-

ing conductive objects in a distance of up to two me-

ters(von Wilmsdorff et al., 2018), while the range of

SENSORNETS 2021 - 10th International Conference on Sensor Networks

capacitive sensing is limited to about 35cm (see 6.2).

A loading mode sensor measures the capacity of a vir-

tual capacitor that is created between electrode and

user, whereas an electric potential sensor measures

the induced current that a user can transmit through

the same virtual capacitor between electrode and him-

self. This also implies that the electric potential sen-

sor of the Linoc toolkit is only able to detect move-

ments, while the loading mode sensor is capable of

sensing non-moving objects. The last difference is the

energy consumption which is lower for electric poten-

tial sensors, but since the microcontroller draws far

more power than all sensors combined and because

the Linoc toolkit is designed to be used with a USB

connection or a USB powerbank, this is not a crucial

point of the toolkit discussion.

4.4 Board Layout

Since one of the main goals in mind while designing

the Linoc prototyping toolkit is fast data acquisition,

the board features numerous connection types. The

Linoc prototyping toolkit can sent data over various

protocols, as well as raw binary data, as shown in Ta-

ble 2.

Table 2: Outputs and their correlating connectors.

Protocoll Connector

Wi-Fi (b/g/n) 2.4GHz antenna

Bluetooth 2.4GHz antenna

USB front board connector

S left and right pinouts

UART micromatch connector

Binary data over GPIO left and right pinouts

Binary data over relais left and right pinouts

Figure 5 clariﬁes the location of all connec-

tors. The prototyping board also has three free pro-

grammable Light-Emitting Diodes (LEDs) as well as

a free programmable button. The second button of the

board is hard wired to reset the microcontroller.

Figure 5: Hardware layout of Linoc.

A mosfet relais was added to board (see Figure 5)

to cover more use cases (see design goal 4) without

the need of further soldering. The relais enables the

user to directly connect LEDs, piezo buzzers or other

hardware that requires more voltage and higher cur-

rents to the board (up to 30V and 1A).

The micromatch connector was added for several

reasons. Since the board should be capable of creating

sensor chains, which was discussed in goal 5, a ribbon

cable is a suitable solution. This is because, again,

no further soldering is required when crimping new

connectors to a ribbon cable. Ribbon cables can be

conﬁgured with any number of connectors which can

be spaced at any distance.

Another important feature of the toolkit is the built

in USB to UART converter. The converter is required

to ﬂash the microcontroller. By integrating it onto

the board an external programmer, common to many

toolkits, is not required here. Note that the connection

over UART/USB does not support live debugging as

JTAG adapters do.

5 SOFTWARE

Central interface to the Linoc toolkit is a console over

the UART bus of the microprocessor, which is con-

nected to the USB interface. Without programming

or ﬂashing, the conﬁguration of sensor groups, sam-

pling frequency and sensor array conﬁguration can be

changed as well as wireless connections established

and system diagnostics be printed. A help command

lists available options. This is usually sufﬁcient to use

the toolkit in own projects and fulﬁlls use case 2, be-

cause no further programming is required.

The software implementation is done with the

ESP-IDF and FreeRTOS in the programming lan-

guage C. This allows advanced users to access low-

level functionality of the microcontroller in order to

modify the software to their needs as aimed for in de-

sign goal 4. Multiple sensors can be connected to a

sensor array (design goal 5) and communicate via I

with each other. After initiating the setup from the

master device (the one connected to the users com-

puter), the order on the array is established by press-

ing a button on the slave devices in the respective or-

der. This allows for various physical setups.

The sensor data is transferred to the host computer

via UART, or transmitted via Wi-Fi, either by posting

to a given TCP server or by hosting a TCP server on

the toolkit to which other entities can connect to (de-

sign goal 1). This data can then be further processed

by the user with her preferred tools or programming

languages (design goal 3).

Linoc: A Prototyping Platform for Capacitive and Passive Electrical Field Sensing

6 EVALUATION

To generate a meaningful and comparable evaluation,

we choose to cover multiple categories identiﬁed in

the meta study by Ledo et al. (Ledo et al., ), in which

evaluation strategies and common pitfalls were iden-

tiﬁed by analyzing 68 published toolkits. Thus we

chose to evaluate in terms of usability, performance

and demonstration.

6.1 User Evaluation

To evaluate usage, a set of tasks was devised for the

participants to explore the functionality of the toolkit.

The instructions besides setup, were vaguely formu-

lated, to determine if the Linoc interfaces are designed

intuitively and the information provided through the

console sufﬁcient. The advanced tasks were only

given as concepts to provide some freedom for the

participants to explore the toolkit. Following the set

of tasks a Likert scale questionnaire was answered by

the participants. The questionnaire is attached in Ap-

pendix 6.3. The study was carried out with eleven

participants. More information about the participants

and what operating systems they used is listed in Ta-

ble 3. Most participants needed some time to familiar-

ize with the toolkit and the way the commands work,

but became more ﬂuent over time. The majority of

participants needed about 20 to 30 minutes to com-

plete the survey.

Table 3: Participant information.

Age 24 - 39, Average: 30

Gender Female: 4, Male: 7

Occupation School: 2, Study: 6, Work: 3

OS used Windows: 5, Mac: 2, Linux: 4

The results of the questionnaire are presented in

Table 4. The usage of the toolkit and even the com-

plex task of setting a sensor array was reported as in-

tuitive. Even participants reporting not experienced in

programming or computer systems mostly did not feel

overwhelmed by the system and a participant feeling

overwhelmed in the beginning was able to ﬁnish the

given tasks with some assistance.

Eight out of the eleven participants reported re-

ceiving assistance during the study. Nearly all partic-

ipants were able to complete all the given tasks suc-

cessfully. In most cases the assistance was limited

to occasional clariﬁcations, only two minor technical

difﬁculties arose during the evaluation which resulted

in receiving more assistance than just clariﬁcations.

As the results of the evaluation in Table 4 show,

the overall feedback is very positive. This is not only

shown by the high marks but also by the low standard

deviation, which indicates that the opinion of all par-

ticipants are nearly the same. Application ideas for

future projects as stated from the users in ”Free ques-

tions” section of the questionnaire centered around

general activity and proximity detection and integra-

tion to cloud computing.

Table 4: Results from the questionnaire, ranging from 1

(strongly agree) to 5 (strongly disagree).

Question

Mean

Median

STD

Experienced

User?

1,91 1 1,6

Similar

systems

already used?

3,09 3 1,51

Tasks

successful?

1,45 1 0,69

Feeling over-

whelmed?

4,18 5 1,25

Usage

intuitive?

1,82 1 1,25

Functionality

sufﬁcient?

1,09 1 0,42

Sensor array

intuitive?

1,82 2 0,98

6.2 Performance

To assess the performance of the loading mode sen-

sors, the Linoc prototyping toolkit was mounted in an

automatic test stand. The test stand consists of a metal

box to shield from external inﬂuences and a conduc-

tive platform that is moved by a stepper motor. It can

be lifted up to 35cm over the sensor.

Figure 6: Data of a test run in the automated test stand in-

cluding the standard deviation.

All sensor values are recorded for several minutes,

then the platform is moved upwards 1cm. For this

test, the pulses of the 555 timer (see 4.2) were counted

for 0,5 seconds, which corresponds to a sensor sam-

SENSORNETS 2021 - 10th International Conference on Sensor Networks

Figure 7: Comparison of the ﬂuid level metering in (Wimmer et al., ) (“sensor reading (units)” is plotted over “Fluid Level

[mm]”) and a reproduction using the same electrode and beer bottle on the right.

pling rate of 2 Hz. This means that over 200 samples

were collected for every position of the movable test

platform.

It is possible to distinguish between two different

distances of the test platform as long as the measured

values don’t overlap. A graphical representation with

error bars was arranged, as illustrated in Figure 6, to

visualize which distances were distinguished success-

fully and thereby concluding the maximal range of the

system. To derive if a distance is distinguishable from

the previous, following metric is used:

¯m

− ¯m

(i−1)

− σ

i−1

> 0

Meaning that if the mean value of a distance ¯m

minus

its standard deviation σ

is still bigger than the mean

value of the previous measured distance ¯m

(i−1)

, in-

cluding its standard deviation σ

i−1

, all sensor values

of these two distances can be clearly assigned.

In our tests, the maximal testable range of 35cm

of the automated test stand was detectable with the

Linoc toolkit, matching the performance of the Open-

CapSense, as shown in Figure 6. Note that this test

stand is shielded from environmental noise so that

the detectable range can be lower in noisy environ-

ments. The big standard deviation, seen at a distance

of 5mm in Figure 6, is the result of a settling time of

the sensor. This is why the ﬁrst values of the load-

ing mode sensor can be observed as more noisy than

values recorded later on.

Another impact factor of the sensor range are cus-

tom ﬁrmwares. A software which generates strong

ﬂuctuations in the energy demand of the processor

will also destabilize the sensors power supply. Very

CPU intensive tasks followed by a CPU idle time can

induce such a behaviour. To prevent this issue, Linoc

was equipped with large decoupling capacitors.

6.3 Demonstration: A Ten Minute Built

of a Fluid Level Metering

CapToolkit (Wimmer et al., ) demonstrated a ﬂuid

level metering for a beer bottle with a 10cm x 10cm

electrode. This demonstration was repeated for com-

parison. Care was taken to buy the same beer brand

(Augustiner Lagerbier Hell). To ensure consitent re-

sults and to exclude the tester, he had to stay at a con-

stant distance and allowing the signal to tune in after

pouring out 4cl beer at a time. A calibration curve

similar to the one presented in (Wimmer et al., ) was

measured and is shown in Figure 7.

Note that the unit on the y-axis differs from the

usual unit throughout this paper. This is due to the

fact, that Wimmer et al. measure the time of the ca-

pacitive charging cycles and not the frequency. The

relationship is t = 1/ f and the absolute value deter-

mined by the sampling length.

REFERENCES

Braun, A. Application and validation of capacitive proxim-

ity sensing systems in smart environments.

Grosse-Puppendahl, T. Capacitive sensing and communica-

tion for ubiquitous interaction and environmental per-

ception.

Grosse-Puppendahl, T., Berghoefer, Y., Braun, A., Wim-

mer, R., and Kuijper, A. OpenCapSense: A rapid pro-

totyping toolkit for pervasive interaction using capac-

itive sensing. In 2013 IEEE International Conference

on Pervasive Computing and Communications (Per-

Com), pages 152–159. IEEE.

Grosse-Puppendahl, T., Braun, A., and Dellangnol, X. Pro-

totyping Capacitive Sensing Applications with Open-

CapSense. 20(2):16–21.

Hamblen, J. O. and van Bekkum, G. M. E. An Embedded

Linoc: A Prototyping Platform for Capacitive and Passive Electrical Field Sensing

Systems Laboratory to Support Rapid Prototyping of

Robotics and the Internet of Things. 56(1):121–128.

Houben, S. and Marquardt, N. WatchConnect: A Toolkit for

Prototyping Smartwatch-Centric Cross-Device Appli-

cations.

Ledo, D., Houben, S., Vermeulen, J., Marquardt, N.,

Oehlberg, L., and Greenberg, S. Evaluation Strate-

gies for HCI Toolkit Research. In Proceedings of the

2018 CHI Conference on Human Factors in Comput-

ing Systems - CHI ’18, pages 1–17. ACM Press.

Marquardt, N., Diaz-Marino, R., Boring, S., and Greenberg,

S. The proximity toolkit: Prototyping proxemic in-

teractions in ubiquitous computing ecologies. In Pro-

ceedings of the 24th Annual ACM Symposium on User

Interface Software and Technology - UIST ’11, page

315. ACM Press.

Savage, V., Zhang, X., and Hartmann, B. Midas: Fabri-

cating custom capacitive touch sensors to prototype

interactive objects. In Proceedings of the 25th An-

nual ACM Symposium on User Interface Software and

Technology - UIST ’12, page 579. ACM Press.

von Wilmsdorff, J., Kirchbuchner, F., Braun, A., and Kui-

jper, A. (2018). Eliminating the ground reference for

wireless electric ﬁeld sensing. In Kameas, A. and

Stathis, K., editors, Ambient Intelligence, pages 90–

99, Cham. Springer International Publishing.

Wimmer, R., Kranz, M., Boring, S., and Schmidt, A. A

Capacitive Sensing Toolkit for Pervasive Activity De-

tection and Recognition. pages 171–180. IEEE.

SENSORNETS 2021 - 10th International Conference on Sensor Networks

Evaluation of the

Prototyping Platform

for Capacitive and

Electrical Field Sensing

1 Introduction

The Linoc toolkit is built for activity detection through electric ﬁeld measurement and has two measurement techniques:

capacitive and passive electric ﬁeld sensing.

Its beneﬁts are easy reconﬁguration of sensor groups and connectivity without knowledge of embedded systems.

2 Setup

The Linoc sensor is connected to the computer via USB. The electrodes are connected to the pin-outs at the side of the

board. Labels are on the bottom side of the chip.

2.1 Connection Setup

To communicate with the sensor a couple of steps are necessary.

Windows:

In the device-manager it can be checked if the Linoc is detected correctly. If this is not the case, the right device driver

has to be installed from www.ftdichip.com. The connection is then made with the program Putty (putty.org).

After the installation, start Putty with the option serial, the com-port from the device-manager, a baud rate 115200 and

without hard- and software control ﬂows. Messages should now appear in the terminal.

Linux:

Start a terminal (Ctrl+Alt+t). The current user has to belong to the group dialout

This can be checked with following command:

groups | grep ’dialout’

If the output is empty the user has to be added to the group using following command:

sudo usermod -aG dialout $USER

A reboot is necessary to apply the changes.

Linux & MacOS Communication happens easiest via Putty.

(if Putty is not available it can be installed with ‘sudo apt install putty’. Alternatives are screen, minicom, or other pro-

grams supporting serial connection). Putty is started with the option serial, the port (like /dev/ttyUSB0) and a baudrate

of 115200

Messages should now appear in the terminal.

� setup was completed successfully � setup was not necessary

� messages from sensor are displayed � setup was not possible

Evaluation of the Prototyping Platform for Capacitive and Passive Electrical Field Sensing

2.2 Visualization (optional)

A graphical visualization helps understanding sensor values. A multitude of programs can be used for this. As example

the Arduino IDE is mentioned here (https://www.arduino.cc/en/Main/Software). The respective port has to be selected

in the menu ‘Tools’. Afterwards the‘Serial Plotter’ can be started from the same menu.

The ‘Serial Monitor’ can be used to change the conﬁguration.

Abbildung 1: Putty window. relevant settings marked with circles

APPENDIX

Questionaire

Linoc: A Prototyping Platform for Capacitive and Passive Electrical Field Sensing

Evaluation of the Prototyping Platform for Capacitive and Passive Electrical Field Sensing

3 Conﬁguration

3.1 Console

The conﬁguration of the Linoc toolkit happens over a console. The console can be reached through pressing the button

on the board, on the side with the connectors notch.

Typing help displays an overview of available commands.

� Console is displayed � Commands are executed

3.2 Changing the Sensor Conﬁguration

With the commands capacitive_switch, eps_switch and frequency_set, the sensor conﬁguration can be changed. With exit

the sensor returns to displaying values. sensor_info displays the current setup.

Please conﬁgure the sensor to the conﬁgurations described in the sections below and check if the values change if you

move your hand towards the measuring electrode. For the conﬁguration it might be necessary to change the placement

of the electrodes on the chip.

3.2.1 Task 1: Conﬁguration #1

Active sensors: Capacitive 1 & 2

Sampling frequency: 2 Hz

� Two value pairs are printed per second

3.2.2 Task 2: Conﬁguration #2

Active sensors: EPS 1

Sampling frequency: 50 Hz

� Values appear noticeably faster

3.3 Task 3: Connect to Wi-Fi

Connect to the provided Wi-Fi hotspot and start a TCP server on the sensor.

Stop the connection again afterwards.

� connection established successfully. � values can be queried from another device

4 Sensor Array

Multiple Linoc sensors can be connected to a sensor array (Daisy Chain).

The command to initiate is ‘daisy_chain init’.

Follow the instructions provided in the console from now on.

The connected devices takes the role of the master and queries sensor values of the other sensors.

4.1 Task 4: Sensor Array Setup

Connect 3 sensors with the provided cable to a sensor array

4.2 Task 5: Conﬁguration

Follow the instructions in the console to setup all sensors.

� Number of sensors detected correctly � Sensor values of all sensors are printed correctly

4.3 Break up the Sensor Array

Stop the sensor array. This is done with the command ‘daisy_chain break’.

Check if the sensor can be used as before.

� Sensor array stopped � Sensor can be used as before

Evaluation of the Prototyping Platform for Capacitive and Passive Electrical Field Sensing

5 Evaluation

Please answer following questions brieﬂy:

Age:

Years

Gender: � F � M � n.b. � n.A.

What is your current occupation: � school � study � apprenticeship � work � something else

Which operating system are you using? � Windows � Mac � Linux � n.A.

Did you receive help with the tasks?: � Yes � No

Are you experienced in programming and/or computer systems?

Have you worked with similar toolkits?

Did you manage to fulﬁll the tasks with the information provided?

Did you at any time had the feeling to be overburdened?

Was the handling of the Linoc toolkit intuitive?

Do you think the functionality is sufﬁcient?

Was the setup of the sensor array intuitive?

5.1 Free questions

What did you like about the system?

What did you dislike?

Could you imagine using the toolkit in a project?

Which applications can you imagine?

Which aspects should be improved?

What would facilitate the handling?

Further comments:

SENSORNETS 2021 - 10th International Conference on Sensor Networks