AEBS and LDWS Exemptions Feasibility Study: 2011 Update. MVWG Meeting, Brussels, 6 th July 2011

Transcription

1 AEBS and LDWS Exemptions Feasibility Study: 2011 Update MVWG Meeting, Brussels, 6 th July 2011

2 Contents Background Method and assumptions Effectiveness estimates Cost estimates Cost Benefit Analyses Results Conclusions and limitations Page 2

3 Background General Safety Regulation aims to mandate AEBS and LDWS for all N2, N3, M2 and M3 vehicles; Exemptions permitted if cost/benefit analysis and/or other vehicle-specific safety aspects show fitment to be inappropriate; TRL completed a detailed cost-benefit analysis of various specific N2/N3 and M2/M3 vehicle types in This update project was to repeat that analysis but to consider the effects of: - Any new evidence on system costs and effectiveness. - Phased implementation proposals for AEBS: FCW/AEBS-M & then AEBS-A. - Draft UNECE performance requirements for LDWS. Page 3

4 Method & Assumptions Short, desk-based study; Information gathered from published literature and stakeholders ( sent to various system suppliers and OEMs), on costs and effectiveness part costs and development costs, and likely changes over time; No updating of EU-27 casualty estimates or vehicle stock/new registration data 2010 study data used; Main analysis limited to N3, N2>7.5t, N2<=7.5t, M3 and M2; LDWS assumed fitted to all new vehicles from Year 1; FCW assumed to be fitted to N2<=7.5t & M2 vehicles, and AEBS-M to N3, N2>7.5t & M3 for Years 1-5. AEBS-A to all vehicles from Year 6 onwards; Cost-benefit calculations based on simple 15 year evaluation period. Page 4

5 Effectiveness estimates LDWS - Draft technical requirements do not require alterations to earlier assumptions about effectiveness; - Stakeholders agreed that, in general, systems conforming to the draft requirements would be similar to existing systems; - No significant differences with capability assumptions used in 2010 study; - Some new evidence on real-world effectiveness, e.g. small field operational trial in USA (overall effectiveness estimate of 26-47%), and a US accident study (relevance in 35-55% of lane departure accidents); study used range of 20-60%, based on various other studies published up to that time. New data suggests a range of 30-50% is now more appropriate. Page 5

6 Effectiveness estimates FCW/AEBS-M/AEBS-A - Stakeholders agreed that FCW systems would be essentially the same as AEBS-M, just without the autonomous braking. Sensors and detection algorithms basically the same; - Stakeholders also suggested that FCW would provide the bulk of the casualty benefit, with AEBS-M and then AEBS-A providing progressively less additional benefit; - Stakeholders and the literature support the view that systems would be more effective with moving targets than with stationary targets; - Literature suggests FCW would prevent 15-25% of all rear shunts. Assumption made that this is equivalent to saving 25-40% of moving target casualties and 5-10% of stationary target; - Model further assumes that AEBS-M adds 10% to these figures, and AEBS-A another 5%. Page 6

7 Effectiveness estimates System Moving target effectiveness Stationary target effectiveness Overall effectiveness FCW 25-40% 5-10% 15-25% AEBS-M 35-45% 15-20% % AEBS-A 40-50% 20-25% % 2010 study used range of 20-50% for both moving target and stationary targets, based on literature available at the time (for moving targets); Stationary target analysis in 2010 assumed a hypothetical future system; 2011 update gives more realistic combined estimates for likely systems. Page 7

8 Cost estimates LDWS - Stakeholders provided initial costs of (range used in 2010 was ); - Range of views on how fast these might change (reduce) over time from no reduction in short term to 20% per annum falls over next five years, due both to higher volumes and innovation; - Other vehicle technologies (e.g. ESC and ABS) were reported to have reduced in cost by 3-12% per annum; - Literature review suggests as a general rule that costs fall by about 20% for every doubling of production volumes; - Information used to define high cost and low cost scenario assumptions. Page 8

9 Cost estimates (LDWS) High scenario 400 Year 1, 3% per annum fall; Low scenario 300 Year 1, 20% per annum fall to Year 5, then no change. Page 9

10 Cost estimates FCW/AEBS - Stakeholders provided initial costs of for moving target systems, assumed to be equivalent to AEBS-M (range used in 2010 was 150-1,000); A supplier suggested Suggested additional costs of per vehicle for nonpneumatic braking systems and/or non-pneumatic rear suspensions; - Agreed that AEBS-M and FCW would be same costs; - Divergence of views on AEBS-A: some thought same as AEBS-M, others thought more expensive for extra system reliability and complexity; - Divergence also on likely cost reductions, from no change in short term to 30% over next five years (equivalent to 7% per annum); - Suggestion also that sensor fusion would mean some LDWS costs would off-set some FCW/AEBS costs; - Agreement that capability more likely to improve, e.g. pedestrians, junctions, for same cost. Page 10

11 Cost estimates FCW/AEBS - Main extra costs (for non-pneumatic vehicles) suggested to be oneoff rather than on-going, e.g: - Real-world driver/braking characteristics data gathering; - Development of a suitable energy reservoir and autonomous brake actuation mechanism. - Other costs for self-levelling device for sensors (vehicles with steel rear suspensions) and possible fitment of full EBS. - Complex set of conditions to factor in, e.g. additional benefits of EBS, proportion of vehicles affected, years to spread development costs; - High scenario assumes 575 Year 1 for FCW/AEBS-M (allowing 25 for sensor fusion), then 3% per annum fall, but back up to 575 in Year 6 for start of AEBS-A; - High also allows 600 per vehicle development costs for 4 years for M2, 300 per vehicle for 4 years for more numerous N2 <=7.5t (assumed to be mostly non-pneumatic braking). Page 11

12 Cost estimates FCW/AEBS - Low scenario assumes 225 Year 1 for FCW/AEBS-M (allowing 50 for sensor fusion), then 7% per annum fall, with AEBS-A costing same as FCW and AEBS-M in Year 6; - Low also allows 500 per vehicle development costs for 2 years for M2, 250 per vehicle for 2 years for more numerous N2 <=7.5t; - Development costs add up to 61million - 146million for M2 and N2<=7.5t; - Costs of self-levelling systems thought by stakeholders to be c. 50 initially, but only applies to small number of vehicles so likely to be already within low-high range used. - Vehicle sub-types use same cost assumptions as relevant major group. N2 categories not split by GVW, e.g. N2 off road, use N2<=7.5t cost model. Page 12

13 Cost Benefit Analyses Phased analysis allowing for effectiveness and gradual introduction into vehicle fleet over 15 year evaluation period (via mandatory fitment to all new vehicles). Results in form of break-even costs (maximum average cost of systems per vehicle to provide a benefit-cost ratio > 1) and benefit:cost ratios (based on estimated system cost ranges). Page 13

14 Results major vehicle types, LDWS LDWS Vehicle Stock (1000s) New reg per year (1000s) Casualty values ( m) Totals BCRs Scenario Lower Mid Upper Year High per vehicle costs ( ) All data in this section below taken from 3% reduction per year TRL 2010 Study (Robinson et al, 2010) Low per vehicle costs ( ) % reduction per year for Years 2-5, then costs stabilise Low Mid High System effectiveness 30% 40% 50% Total vehicles fitted (1000s) N High costs ( m) Low Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) Total vehicles fitted (1000s) N2>7.5t High costs ( m) Low Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) Total vehicles fitted (1000s) N2<=7.5t High costs ( m) Low Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) Total vehicles fitted (1000s) M High costs ( m) Low Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) Total vehicles fitted (1000s) M High costs ( m) Low Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) Page 14

15 Results major vehicle types, AEBS AEBS Vehicle Stock (1000s) New reg per year (1000s) Casualty values ( m) Totals BCRs Scenario Lower Mid Upper Year High per vehicle costs ( ) All data in this section below taken from 3% reduction per year FCW/AEBS-M AEBS-A TRL 2010 Study (Robinson et al, 2010) Low per vehicle costs ( ) % reduction per year FCW, AEBS-M & AEBS-A all at same cost FCW Low Mid High AEBS-M Low Mid High AEBS-A Low Mid High Effectiveness - moving 25% 32.5% 40% 35% 40% 45% 40% 45% 50% Effectiveness - stationary 5% 7.5% 10% 15% 17.5% 20% 20% 22.5% 25% Total vehicles fitted (1000s) High per vehicle cost supplement ( ) Low per vehicle cost supplement ( ) N Moving High costs ( m) Low Stationary Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) AEBS-M AEBS-A Total vehicles fitted (1000s) High per vehicle cost supplement ( ) Low per vehicle cost supplement ( ) N2>7.5t Moving High costs ( m) Low Stationary Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) AEBS-M AEBS-A Total vehicles fitted (1000s) High per vehicle cost supplement ( ) Low per vehicle cost supplement ( ) N2<=7.5t Moving High costs ( m) Low Stationary Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) FCW AEBS-A Total vehicles fitted (1000s) High per vehicle cost supplement ( ) Low per vehicle cost supplement ( ) M Moving High costs ( m) Low Stationary Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Mid costs ( m) Mid benefits ( m) AEBS-M AEBS-A Total vehicles fitted (1000s) High per vehicle cost supplement ( ) Low per vehicle cost supplement ( ) M Moving High costs ( m) Low Stationary Low costs ( m) Mid High benefits ( m) High Low benefits ( m) Page 15 Mid costs ( m) Mid benefits ( m) FCW AEBS-A

16 Results - LDWS Vehicle Type Page 16 Stock (1000s) New registrations per annum (1000s) Effectiveness System costs (Year 1) EU-27 annual estimates LDWS Target population (Casualties) Phased Break-Even Costs ( ) (15 Yrs, fit to all new vehicles) Note 1 Updated Phased Benefit:Cost Ratios (15 Yrs, fit to all new vehicles) 2010 Benefit-Cost Ratios (from Robinson et al, 2010) Lower Mid Upper Lower Mid Upper Lower Mid Upper Lower Mid Upper 30% 40% 50% N ,289 1, N3 >16t ,295 1, N3 >16t Rigid (not towing) , N2<=7.5t Not calculated N2 >5 <=7.5t N2 >7.5t ,230 1, N3 >16t Off Road N2 <=5t , N2 Off Road , N3 >16t >3 axles N3 <=16t , N3 <=16t Rigid (not towing) , N3 >16t Special Purpose N2 Special Purpose N3 <=16t Special Purpose ,534 2,617 4, N3 <=16t Off Road ,906 2, N2 >3 axles * ,155 4,981 7, N3 <=16t >3 axles * ,833 5, N2 Tractor unit <=7.5t * M ,823 2,988 4, M3 Class I/II/III ,005 3,288 4, M M3 articulated * ,007 9,868 13, M3 Class A M3 >3 axles * ,588 45, * NB. The numbers of these vehicles in use in GB are very low (data from other countries not available), so very low numbers of casualties are to be expected - EU- 27 estimates are thus subject to significant uncertainty. Note 1 Phased break-even costs here are equivalent to the average per vehicle cost over the 15 year period.

17 Results AEBS Vehicle Type Years 1-5 system Stock (1000s) New registrations per annum (1000s) Effectiveness System costs (Year 1) EU-27 annual estimates AEBS Target population (Casualties) Phased Break-Even Costs ( ) Phased Benefit:Cost Ratios (All Rear Shunts) (15 Yrs, fit to all new vehicles) Note 1 (15 Yrs, fit to all new vehicles) Lower Mid Upper Lower Mid Upper Lower Mid Upper Lower Mid Upper FCW Moving/Stationary 25%/5% 32.5%/7.5% 40%/10% AEBS-M Moving/Stationary 35%/15% 40%/17.5% 45%/20% AEBS-A Moving/Stationary 40%/20% 45%/22.5% 50%/25% Benefit-Cost Ratios (from Robinson et al, 2010) (Non-Stationary Targets) N3 AEBS-M ,138 1,708 2, N3 >16t AEBS-M ,110 1,666 2, N3 >16t Rigid (not towing) AEBS-M ,062 1, N2<=7.5t FCW Not calculated N2 >5 <=7.5t FCW N3 >16t >3 axles AEBS-M ,292 1,940 2, N3 >16t Off Road AEBS-M ,117 1,677 2, N3 <=16t AEBS-M ,501 2,255 3, N3 <=16t Rigid (not towing) AEBS-M ,392 2,085 2, N2 >7.5t AEBS-M , N2 <=5t FCW N3 >16t Special Purpose AEBS-M N2 Special Purpose FCW , N2 Off Road FCW N3 <=16t Special Purpose AEBS-M ,848 4,308 6, N3 <=16t Off Road AEBS-M ,285 1, N2 >3 axles * FCW ,013 1, N3 <=16t >3 axles * AEBS-M ,540 3, N2 Tractor unit <=7.5t * FCW M3 AEBS-M ,276 1, M3 Class I/II/III AEBS-M ,407 2, M2 FCW M3 Class A AEBS-M M3 articulated * AEBS-M M3 >3 axles * AEBS-M ,575 2, * NB. The numbers of these vehicles in use in GB are very low (data from other countries not available), so very low numbers of casualties are to be expected - EU-27 estimates are thus subject to significant uncertainty. Note 1 Phased break-even costs here are equivalent to the average per vehicle cost over the 15 year period. Page 17

18 Conclusions & Limitations Decisions regarding the applicability of the cost-benefit analysis and technical evaluations to questions of possible exemption for specific vehicle types are a political matter for stakeholders and legislators to debate and decide upon; Important, though, to remember that the technical assessments, benefit-cost ratios and break-even costs presented in this report are all based on a wide variety of assumptions and subject to limitations, e.g: - Past accident/casualty statistics are imperfect predictors of future patterns. - Accident analyses based on samples of data from, at best, 3 Member States and often only 1 or 2, giving a high level of uncertainty when making EU27 estimates, particular where low numbers of vehicles and accidents combine. - The vehicle types assessed within the accident databases are, generally speaking, not exact matches to the vehicle types proposed by stakeholders for exemptions. - Different AEBS and LDWS architectures will have different operational characteristics, costs and effectiveness in differing applications and accident scenarios. The analyses presented here are entirely generic and can only attempt to allow for these variations by using quite wide ranges of likely effectiveness and costs. Page 18

19 THANK YOU! Brian Robinson, Safety Division Page 19