The ADAS SWOT Analysis

A Strategy for Reducing Costs and Increasing Quality in ADAS Testing

Andreas Haja, Carsten Koch and Lars Klitzke

Faculty of Technology, Hochschule Emden/Leer, University of Applied Sciences, Emden, Germany

Keywords: Testing, Validation, ADAS, Advanced Driver Assistance Systems, Autonomous Vehicles, Test Strategy,

SWOT, Design of Experiments, Sensors, Automotive.

Abstract: In a remarkably short time, advanced driver assistance systems (ADAS) have become a major driver of

innovation in the auto industry: It is expected that autonomous vehicles will profoundly change the very

definition of mobility. In addition to mastering technical challenges, increasing automation requires a

significant amount of testing and thus a huge investment in test resources. This poses a serious cost factor for

existing companies and a high entry barrier for new market entrants. In addition, strong demand for engineers

worldwide also makes it difficult to allocate sufficient manpower. Consequently, tests are often performed by

teams with limited experience and high staff turnover. To reduce test duration while ensuring high levels of

quality and a focus on the most relevant aspects, this paper presents a new method for creating efficient test

strategies which builds on the well-known SWOT analysis and extends its use to ADAS-related scenarios.

The ADAS SWOT analysis provides a structured process which facilitates the identification of risks and

opportunities associated with new technology and assesses its impact on ADAS products from a customer

perspective. The method has been tailored to fit the needs of research and advance development and helps

increase both product quality and time-to-market.

1 INTRODUCTION

Automakers and suppliers are currently competing

fiercely to be among the first to launch a fully

autonomous driving solution into the market. In

addition to solving technical challenges, the high

amount of testing required for such systems poses a

significant cost factor for existing players and a high

entry barrier for newcomers to the ADAS segment.

This paper proposes a method for reducing cost and

increasing quality of testing, especially in corporate

research and in advance development teams.

A typical ADAS development chain consists of

the stages research, advance development and series

development. Each stage is associated with different

test efforts that accompany the engineering tasks.

One of the main tasks of the research stage is the

identification of new technology components and the

assessment of their potential for creating novel

customer functionality. The focus in this stage is on

constructing and showcasing a working prototype

while the effort invested into testing is often limited.

During advance development, the prototype is

tested thoroughly under a multitude of conditions,

including adverse scenarios such as heavy weather

and complex traffic situations. Also, the potential for

cost-savings is investigated, e.g. by replacing costly

sensors or computationally expensive algorithms.

The focus in this stage is on verifying the industrial

feasibility in terms of cost, package and robustness.

With development progressing towards series

production, test efforts increase exponentially. It is

therefore important to identify the major threats to a

new product as well as its potential opportunities at

an early development stage. To achieve this goal, a

unified test strategy is required which stretches along

the entire development chain and which helps the

engineers to decide on whether to move on to the next

stage or to freeze or even abandon a project.

In practice, ADAS test design is often based on

expert knowledge and on engineering intuition.

Although practical in many situations, this approach

presumes the existence of seasoned experts on the

team who can leverage their extensive knowledge to

devise solid test cases. If experts are not available

320

Haja, A., Koch, C. and Klitzke, L.

The ADAS SWOT Analysis - A Strategy for Reducing Costs and Increasing Quality in ADAS Testing.

DOI: 10.5220/0006354103200325

In Proceedings of the 3rd International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2017), pages 320-325

ISBN: 978-989-758-242-4

however, test design becomes a challenging task,

especially for newly assembled teams. In such cases,

a structured and easily applicable process is required

which helps the team to develop a test strategy that

fits the demands of the development project, such as

the construction of a new ADAS prototype.

An overview of current approaches to ADAS

testing has been presented in (Stellet et al., 2015). The

majority of publications contains proposals for either

performance metrics (Smith, 2011 and Fritsch, 2013)

or for testing a specific sensor or function (Fabris,

2014). However, to the best knowledge of the authors,

a generic and structured method for ADAS test

strategy development has not yet been published.

The method proposed in this paper builds on the

well-known SWOT

analysis, which is often used for

assessing the pros and cons of complex decisions. The

SWOT analysis helps companies and institutions to

benefit from market opportunities and to deflect risks

by leveraging internal strengths and by compensating

known weaknesses. In its original form, strengths (S)

and weaknesses (W) look at internal factors of a

product or business venture, such as unique features,

or financial indicators, whereas opportunities (O) and

threats (T) look at external factors such as

competitors, new technology or market trends.

Within the ADAS space, the SWOT analysis has

been previously mentioned in several publications

such as in (Blythe, 2002) or in (Diakaki, 2015).

However, none of these publications addresses the

topics of ADAS development or testing.

To meet the requirements of the individual stages

of the development chain, a modified version of the

original SWOT analysis is presented in this paper. In

the proposed ADAS SWOT analysis, the internal

perspective (strengths and weaknesses) is directed to

focus on the technological properties of a specific

component (e.g. a sensor) whereas the external

perspective (opportunities and threats) looks at the

entire system from a product-oriented viewpoint.

The ADAS SWOT analysis is intended to help

OEMs or Tier-1s to benefit from new technology

components that become available and to deflect risks

that might arise from the use of these components.

The new method can be applied in two scenarios

common in both research and advance development:

1. Assess whether a new ADAS product should be

developed using a specific technological setup.

2. Decide whether a component within an ADAS

product should be replaced by an alternative.

Strengths, Weaknesses, Opportunities and Threats

Throughout this paper, the second scenario is used

to illustrate the new method. However, ADAS SWOT

can be applied to the first scenario in a similar way.

This paper is structured as follows: In Section 2,

the ADAS SWOT analysis is introduced. In Section

3, the method is applied to a well-known scenario to

illustrate the benefits. In Section 4, major findings are

summarized and an outlook on future work is given.

Figure 1: Stages of the ADAS SWOT analysis.

2 THE ADAS SWOT ANALYSIS

The proposed ADAS SWOT analysis consists of four

consecutive steps (see Figure 1):

1. Assess technology component: Identify major

strengths and weaknesses of a new component.

2. Assess ADAS product: Identify opportunities

and threats which show the impact of strengths

and weaknesses from (1) on the ADAS product.

3. Rank and prioritize: Rank and prioritize the

findings from (2) regarding safety and quality.

4. Test-case definition: Define a set of tests to

verify whether the findings from (3) exhibit the

predicted quality and safety increase or decrease.

The tests in (4) will support the decision on whether

to select the new component for the ADAS product.

The ADAS SWOT Analysis - A Strategy for Reducing Costs and Increasing Quality in ADAS Testing

321

2.1 Technology Assessment

In the first step of the ADAS SWOT analysis, only

strengths and weaknesses of the new component are

discussed without addressing the ADAS product in

which the component is to be integrated yet.

The aim of the assessment is to reliably identify

all major advantages and technology highlights as

well as disadvantages and potentially hazardous

properties from an engineering perspective.

The analysis is conducted by the development

team. To improve results, an expert with prior

experience with similar technology and its use in an

automotive environment should support the team.

Questions to be answered are:

 Strengths (S)

S1. What are the major assets of the new component?

S2. What makes the new component better than

alternative solutions?

S3. Under which conditions does the new component

perform well?

 Weaknesses (W)

W1. What are the major areas of concern with the new

component?

W2. In which areas do alternatives perform better?

W3. Under which conditions does the new component

perform poorly?

The result of the first step (see

Figure 1, step 1) is

a list of answers to the strength-related (S

, S

) and

to the weakness-related questions (W

, W

2.2 Product Assessment

In the second step of the ADAS SWOT analysis, the

team focus is directed towards the ADAS product. In

addition to the engineering perspective from the

previous step, a product-centric view ensures that the

team focusses on the implications of technological

strengths and weaknesses for the customer with

regard to safety and quality.

In step 2a of

Figure 1, the team should call upon

the support of the ADAS functional manager and of

the product management to compile a list of use-cases

U, which reflect the expected behavior of the ADAS

product in scenarios with high customer-relevance.

In step 2b of

Figure 1, the team must discuss the

implications of all strengths and weaknesses from

step 1 for each use-case from step 2a. Improvements

Failure Mode and Effects Analysis

to or expansions of the existing features of the ADAS

product are termed opportunities in this context

whereas feature degradations are termed threats.

Questions to be answered for each use-case are:

 Opportunities (O)

O1. How does the new component improve quality?

O2. How does the new component expand features?

 Threats (T)

T1. How does the new component degrade quality?

T2. How does the new component reduce features?

All answers must relate to the strengths in S

, S

and S

and to the weaknesses in W

, W

and W

as well

as to specific use-cases in U.

The result of the second step (see

Figure 1, step 2)

is a list of answers to both the opportunity-related

) and the threat-related questions (T

, T

2.3 Use-Case Ranking

In the third step of the ADAS SWOT analysis, a

ranking scheme is applied to all opportunities and

threats which assesses the impact of the new

component on both safety and quality for each use-

case by attributing individual safety and quality

measures. The goal is to determine the use-case

relevance for the test-case definition in Section 2.4.

The measures are divided into four categories:

 Safety Decrease: Rates how a degraded use-case

reduces occupant or road user safety.

 Safety Increase: Rates how an improved use-

case increases occupant or road user safety.

 Quality Decrease: Rates how a degraded or

removed use-case negatively affects customers.

 Quality Increase: Rates how an improved or

expanded use-case positively affects customers.

Safety decrease assessment in Section 3a is

similar to a FMEA

. For all threats and the associated

use-cases, the following parameters are evaluated :

 Probability of exposure P

E,S-

: How probable is it

that the vehicle within its life time will be

exposed to a situation where the safety decrease

might lead to an accident?

(1 : very unlikely — 5 : very likely)

 Severity S : How severe would be the

consequences of an accident for occupants and/or

road users?

VEHITS 2017 - 3rd International Conference on Vehicle Technology and Intelligent Transport Systems

322

(1 : negligible — 5 : severe injuries / death)

 Detectability D : Can the system detect an

increased probability for threat occurrence?

(1 : very likely — 5 : very unlikely)

The ranking for each threat and its associated use-

case

with regard to safety decrease is computed as:







,

⋅⋅

(1)

Although the definition of each parameter differs, the

method of computation for safety increase (Section

3a), quality increase and quality decrease (Section 3b)

is identical:

 Probability of exposure

o Safety increase (P

E,S+

): How probable is it

that the vehicle within its life time will be

exposed to a situation where the safety

increase might avoid an accident?

o Quality increase (P

E,Q+

) / decrease (P

E,Q-

How probable is it that the quality increase /

decrease will be experienced daily?

(1 : very unlikely — 5 : very likely)

 Impact

o Safety increase (I

S+

): How significant is the

positive impact of the safety increase on

occupant and/or road user safety?

o Quality increase (I

Q+

) / decrease (I

Q-

): How

significant is the impact of the quality

increase / decrease on customer satisfaction?

(1 : negligible — 5 : profound impact)

Probability of exposure and impact are computed as:







,

⋅



(2)







,

⋅





(3)







,

⋅





(4)

Note that opportunities or threats may result in

different rankings if the use-case is changed.

The result of the third step (see

Figure 1, step 3) is

a list of threats and opportunities ranked according to

the impact on safety and quality. The ranking is the

basis for the test-case prioritization in the next step.

2.4 Test-Case Definition

In the fourth and final step of the ADAS SWOT

analysis, a set of test cases is defined. All tests will be

designed to verify the strengths and weaknesses

identified in Section 2.1 during the technology

assessment. By propagating the test results through

the ADAS SWOT analysis, the validity of the

opportunities and threats and thus the influence of the

new component on the ADAS product is verified.

Categories to define each test scenario (based on

the use-cases selected in the second step) are:

 Expectation: Expected behaviour of the new

component leading to an opportunity or threat.

 Test setup: Behaviour of prototype vehicle and

other test participants (e.g. lead car)

 Test parameters: Variables to change between

scenarios to assess their influence on test results.

 Performance assessment: Measurable criterion

for the assessment of test performance.

In practice, categories will often differ depending

on the system being tested and on the test context. For

example, a description of categories for emergency

braking can be found in (EUNCAP, 2015).

By linking technical properties with product

features, a meaningful and focussed test set is created

with the potential to reduce test efforts significantly.

Also, the ranking scheme applied in the previous

section ensures that the most relevant test cases are

executed first. This allows for a firm decision on the

new component in an early stage in case of a

confirmed hazardous threat or if an expected

significant opportunity did not manifest.

3 ADAS SWOT EXAMPLE

In this section, the ADAS SWOT analysis is used to

assert whether a mono camera sensor could be a

replacement for a radar sensor in an adaptive cruise

control system (ACC). This question has already been

discussed extensively such as in (Stein et al., 2003),

(Dagan, 2004) or (Ingle, 2016).

From a commercial perspective, the use of a

camera instead of a radar sensor offers the potential

to reduce the price of ACC and thus increase the

number of car models in which it may be integrated.

Also, additional customer value can be created by

using the camera for lane or traffic sign detection.

This section demonstrates the ADAS SWOT

principles and has been shortened for brevity.

3.1 Technology Assessment

In the first step, the goal is to identify strengths and

weaknesses of the camera. Answers to questions S

1-3

and W

1-3

from Section 2.1 are shown in Table 1.

The ADAS SWOT Analysis - A Strategy for Reducing Costs and Increasing Quality in ADAS Testing

323

Table 1: Assessment of strengths and weaknesses.

Step 1: Camera technology assessment

Strengths

Assets of the camera sensor?

 lateral information on vehicles

Major advantages over radar?

 large opening angle

 detects lane markings, speed signs,etc.

 can measure road surface structure

Conditions with increased performance?

 daylight and cloudy sky

Weaknesses

Drawbacks of the camera sensor?

 reduced performance with low contrast and

overexposure

Major disadvantages compared to radar?

 short detection range

 low range accuracy

Conditions with decreased performance?

 dusk, darkness

 heavy weather (snow, rain, fog, dust)

 direct sunlight

Table 2: Assessment of use-cases, opportunities, threats.

Step 2: ACC product assessment

1 ACC cruising

Opportunities

: adapt to speed signs and road quality (S

, S

)

Threats

: reduced max. speed (W

12,

, W

)

2 Target following

Opportunities

: reduced FP/FN on curvy roads (S

)

Threats

: jittery distance control (W

, W

)

: inadvertent acceleration (W

22,

)

3 Cutting vehicles

Opportunities

: fast reaction to cutting vehicles (S

, S

)

Table 3: Safety (R

S+

) and quality increase (R

Q+

) for

opportunities.

Step 3a: Opportunity ranking

I R

S+

Q+

Prio

4 3 12 3

5 2 10 4

5 4 20 1

Table 4: Safety (R

S-

) and quality decrease (R

Q-

) for threats.

Step 3b: Threat ranking

S D I R

S-

Q-

Prio

5 3 15 3

4 3 12 4

4 4 4 64 1

Table 5: Test scenarios based on opportunities and threats.

Step 4: Test scenarios

Negative test scenarios (excerpt)

ID Pri Test description

Expectation:

 Loss of lead car track (false negative)

results in inadvertent acceleration

Test setup:

 Prototype follows lead car on highway

Test parameters:

 following distance (near – far)

 ambient light (day, dusk, night)

 weather (sun, cloudy, rain, fog)

Performance assessment:

 lead car track stability

Positive test scenarios (excerpt)

ID Pri Description

Expectations:

 Wider opening angle enables early

detection of vehicles cutting into

prototype lane

Test setup:

 Vehicle from neighbouring lane cuts

into prototype lane on highway

Test parameters:

 Lead car distance (near – far)

 ambient light (day, dusk, night)

 weather (sun, cloudy, rain, fog)

Performance assessment:

 time until stable track of lead car

VEHITS 2017 - 3rd International Conference on Vehicle Technology and Intelligent Transport Systems

324

4 SUMMARY AND OUTLOOK

In this paper, a novel method for designing efficient

tests for ADAS and autonomous vehicles has been

proposed. The main advantage of the method is the

inherent consistency of the test strategy which

progresses from the technological properties of a

system component towards its implications for the

final ADAS product from a customer perspective.

The method has been designed to meet the needs

of newly assembled teams which must achieve high-

quality test results in a limited amount of time.

Using a well-known example from recent ADAS

history, the replacement of a radar sensor with a mono

camera for ACC has been chosen as an example to

illustrate the use of the ADAS SWOT analysis.

In future work, the following issues will be

addressed to further optimize and improve the testing

process:

 Apply ADAS SWOT to new sensor technologies

(e.g. solid-state LiDAR, stereo camera).

 Expand ADAS SWOT to meet the needs of

autonomous vehicle development.

 Further reduce the required amount of expert

knowledge by partially automating the testing

process (e.g. by defining test criteria and

environmental parameters based on ontologies).

REFERENCES

J.E. Stellet, M.R. Zofka, J. Schumacher, T. Schamm,

F. Niewels and J.M. Zöllner. Testing of Advanced

Driver Assistance Towards Automated Driving: A

Survey and Taxonomy on Existing Approaches and

Open Questions, IEEE 18th International Conference

on Intelligent Transportation Systems, 2015.

K. Smith, R. Schweiger, W. Ritter, and J.-E. Kallhammer,

Development and evaluation of a performance metric

for image-based driver assistance systems, IEEE

Intelligent Vehicles Symposium , 2011. 

J. Fritsch, T. Kuhnl, and A. Geiger, A new performance

measure and evaluation benchmark for road detection

algorithms, 16th International IEEE Conference on

Intelligent Transportation Systems, 2013. 

S. Fabris, J. D. Miller, and J. Luo, Validation of an AEB

system, 3rd International Symposium on Road Vehicles

Functional Safety Standards and Its Application, 2014.

P.T. Blythe, Can ITS Satisfy the Demands of the UK

Integrated Transport White Paper and Subsequent 10

Year Transport Plan: A SWOT Analysis. Proceedings of

the 9th World Congress on Intelligent Transportation

Systems, 2002.

C. Diakaki, M. Papageorgiou, I. Papamichail and I.

Nikolos, Overview and Analysis of Vehicle Automation

and Communication Systems from a Motorway Traffic

Management Perspective, Transportation Research Part

A: Policy and Practice, Volume 75, 2015.

G. P. Stein, O. Mano and A. Shashua, Vision-based ACC

with a single camera: bounds on range and range rate

accuracy, IEEE Intelligent Vehicles Symposium, 2003.

E. Dagan, O. Mano, G. P. Stein and A. Shashua, Forward

collision warning with a single camera, IEEE

Intelligent Vehicles Symposium, 2004, pp. 37-42.

S. Ingle, M. Phute, Tesla Autopilot : Semi Autonomous

Driving, an Uptick for Future Autonomy, International

Research Journal of Engineering and Technology,

Volume 3, Issue 9, 2016.

EUNCAP (European New Car Assessment Programme),

Test protocol - AEB systems, EuroNCAP, Test Protocol

1.1, June 2015.

S. Geyer, M. Baltzer, B. Franz and S. Hakuli, Concept

and.development of a unified ontology for generating

test and use-case catalogues for assisted and automated

vehicle guidance,” Intelligent Transport Systems, IET,

vol. 8, no. 3, pp. 183–189, 2014.

The ADAS SWOT Analysis - A Strategy for Reducing Costs and Increasing Quality in ADAS Testing

325