633
cle to a standstill. This requires that the CMS vendor and the OEM must agree on a safe L9
intervention strategy and is crucial to the success of CMS in practice. A problem encountered
with this approach is reducing the machine speed to this pre-defined safe brake speed. There
are multiple approaches employed by OEMs to achieve this, chief among which is de-rating
or de-throttling the power unit. This introduces a multitude of variables that influence the
distance required to reach the safe braking speed from the operating speed. These variables
include the grade of the road the TMM is travelling on, the payload of the TMM, gear selection of the TMM and the driveline type. For the lab-scale tests, the type of intervention above
and below the safe braking speed is agreed upon a priori in the form of an acceleration set
point, with the test vehicles then matching this performance using the brake robot. For the
single machine tests on actual TMMs, the CMS vendor needs to take the variance in braking
performance into account.
Nevertheless, the success criteria for L9 intervention can be summarised as follows:
• The CMS may not generate a L9 Stop signal above the safe braking speed.
• The LO must slow down to the safe braking speed at a distance from the RO not less than
the distance at which the L9 Stop signal is expected.
• Once the LO and RO are stationary, the gap between the two objects must be greater than
a specified value.
For the majority of the test configurations, the distance is measured between the two test
vehicles/machines. Due to the complexity associated with the intersection test configurations,
the distance is measured in these cases to the anticipated intersection point. Because of the
variability when performing intersection tests, the measurement tolerance is increased for
these test runs. Typical measurement tolerance is ± 0.5 m which is then increased to ± 2 m for
intersection tests.
2.4.2 False positive criteria
The false positive success criteria differs from the false negative criteria, because in the false
positive tests the triggering of warnings or interventions is undesired. False positive warnings
or interventions may hamper production and have a negative impact on the mining activity.
Test configurations that involve false positive tests are Test Configurations 6, 7, 9, and 12. In
these test configurations, the objects are separated an agreed upon distance and the tests then
conducted. The triggering of control levels when the objects are not on a collision course will
then be registered as failed test runs.
2.5 Reporting
The final step in the test methodology is reporting on the results. This is of the utmost importance, as this provides the CMS vendor with a thorough report on the performance of the
system and it may be used by mines to compare the CMS capability with their requirement.
The report contains a description of the scope of the tests (the scope is nominated by the
CMS vendor prior to testing) and the actual tested capability. An example of the scope is
shown in Figure 1 and tested capability is shown in Figure 6. The report thus gives the reader
insight into the maturity of the CMS solution and the vendor’s own understanding of the
capabilities and limitations of the solution.
3 SAMPLE TEST RESULTS
Figure 7 shows an example of the test results for Test Configuration 3 – Head-head (Straight)
and Figure 8 shows an example of the test results for Test Configuration 1 – Crossing (90°).
The two objects’ speeds are shown in the ‘a) Object speed’ plots and the objects’ positions are
shown in the ‘b) Object position’ plots. Also shown in these figures are the positions where
the different control levels were triggered.
cle to a standstill. This requires that the CMS vendor and the OEM must agree on a safe L9
intervention strategy and is crucial to the success of CMS in practice. A problem encountered
with this approach is reducing the machine speed to this pre-defined safe brake speed. There
are multiple approaches employed by OEMs to achieve this, chief among which is de-rating
or de-throttling the power unit. This introduces a multitude of variables that influence the
distance required to reach the safe braking speed from the operating speed. These variables
include the grade of the road the TMM is travelling on, the payload of the TMM, gear selection of the TMM and the driveline type. For the lab-scale tests, the type of intervention above
and below the safe braking speed is agreed upon a priori in the form of an acceleration set
point, with the test vehicles then matching this performance using the brake robot. For the
single machine tests on actual TMMs, the CMS vendor needs to take the variance in braking
performance into account.
Nevertheless, the success criteria for L9 intervention can be summarised as follows:
• The CMS may not generate a L9 Stop signal above the safe braking speed.
• The LO must slow down to the safe braking speed at a distance from the RO not less than
the distance at which the L9 Stop signal is expected.
• Once the LO and RO are stationary, the gap between the two objects must be greater than
a specified value.
For the majority of the test configurations, the distance is measured between the two test
vehicles/machines. Due to the complexity associated with the intersection test configurations,
the distance is measured in these cases to the anticipated intersection point. Because of the
variability when performing intersection tests, the measurement tolerance is increased for
these test runs. Typical measurement tolerance is ± 0.5 m which is then increased to ± 2 m for
intersection tests.
2.4.2 False positive criteria
The false positive success criteria differs from the false negative criteria, because in the false
positive tests the triggering of warnings or interventions is undesired. False positive warnings
or interventions may hamper production and have a negative impact on the mining activity.
Test configurations that involve false positive tests are Test Configurations 6, 7, 9, and 12. In
these test configurations, the objects are separated an agreed upon distance and the tests then
conducted. The triggering of control levels when the objects are not on a collision course will
then be registered as failed test runs.
2.5 Reporting
The final step in the test methodology is reporting on the results. This is of the utmost importance, as this provides the CMS vendor with a thorough report on the performance of the
system and it may be used by mines to compare the CMS capability with their requirement.
The report contains a description of the scope of the tests (the scope is nominated by the
CMS vendor prior to testing) and the actual tested capability. An example of the scope is
shown in Figure 1 and tested capability is shown in Figure 6. The report thus gives the reader
insight into the maturity of the CMS solution and the vendor’s own understanding of the
capabilities and limitations of the solution.
3 SAMPLE TEST RESULTS
Figure 7 shows an example of the test results for Test Configuration 3 – Head-head (Straight)
and Figure 8 shows an example of the test results for Test Configuration 1 – Crossing (90°).
The two objects’ speeds are shown in the ‘a) Object speed’ plots and the objects’ positions are
shown in the ‘b) Object position’ plots. Also shown in these figures are the positions where
the different control levels were triggered.