MANTA Summary Report External Stuart Rowell Principal - - PowerPoint PPT Presentation

MANTA

Summary Report – External

Stuart Rowell

Principal Technologist – Systems (Automation & Safety)

Connected Places

Outline

Urban cities face growing mobility challenges through increased density of land use and higher levels of traffic flow. This can lead to increased journey times, higher levels of pollution, and less effective use of available space for residential and commercial purposes. Improving mobility through enabling new travel solutions and supporting innovative solutions is at the heart of what the Connected Places Catapult do.

Connected Places

Outline

Coastal cities such as Southampton face significant mobility challenges due to the geography

• f the land, which naturally funnels traffic down a limited number of arterial routes and

highways, together with periodically high traffic densities as a result of port activities. In addition to the direct impact on road transport, secondary impacts to the area can be felt through under utilisation, under development and under investment of areas which may be geographically close, but not well connected to the commercial center.

Connected Places

Outline

The Connected Places Catapult identified the possibility to improve transport through innovative water-based solutions. While modelled

• n the use case of the City of Southampton, the

technology could equally be applied to other cities facing transport challenges as a result of large bodies of water creating a natural barrier. Innovative Autonomous Water Taxis (AWTs) could provide an on-demand environmentally friendly service connecting travellers to a number of key

• destinations. The service could operate all day,

throughout the year, alleviating some of the road- based travel and making existing under-developed areas more attractive for both residential and commercial use.

Connected Places

Current study and purpose of this document

The Connected Places Catapult (CPC) have conducted a feasibility study into a new passenger service in the Southampton area. This would be a passenger- carrying autonomous ferry taxi service, possibly incorporating an on-demand, app-based user interface. The concept of this service is based on the city of Southampton and the Solent area of water. The city centre

• f Southampton is a central business district, and the Port
• f Southampton (west of the city center) a busy commercial
• port. Road-based transport from surrounding areas to both

areas is a significant problem arising from the geography of land and water. The port of Southampton is also a busy port for commercial and cargo vessels travelling up and down the Solent between the port and the English Channel, which raises a particular challenge for operating such a service.

Ocean Village Woolston Town Quay Hythe Warsash Fawley Marchwood Hamble Point

Connected Places

Current study and purpose of this document

The city of Southampton and the Solent area of water was chosen as a model example of where this transport solution could bring real benefits, and equally because of the challenges that marine autonomy would face if operating in this busy area. However, the project would look generically at the challenges and opportunities of the concept as a more widely exploitable technology.

• Demand Modelling: Conducting

interviews and focus groups to understand the willingness of potential users of such a service, and draw out some key user requirements

• Business Modelling:

Understanding the commercial viability of the service, including the trade-offs between size of boat, frequency of operation and likely ticketing costs

• Safety: Passenger safety is paramount,

and a wealth of regulations and processes are in place to ensure this. The current barriers need to be understood, together with new safety hazards that would need to be addressed The study looked at the concept in three distinct areas:

Connected Places

Outline

Within the feasibility project, sea trials were conducted to gather data in support of the project assumptions.

Connected Places

Demand Modelling

Identifying user needs: Focus groups

Connected Places

Date Location Area* Age 29th October 2018 Totton West 18-44 29th October 2018 Totton West 45+ 30th October 2018 Woolston East 18-44 30th October 2018 Woolston East 45+

To help identify the broad needs of travellers into and out of Southampton, four focus groups were conducted in total: two groups resided on the West side of the water and two groups resided on the East side, each with a spread of ages from each location.

Note: see map on page 3

The participants were asked about their current journeys in and around Southampton, what they liked and didn’t like about local travel, their opinions on autonomous vehicles in general and then specifically boats, and how they would feel about an autonomous ferry service on Southampton Water.

Identifying user needs: Telephone interviews

West Central East Postcode Interviews Postcode Interviews Postcode Interviews SO40 57 SO14 32 SO19 56 SO45 43 SO15 28 SO31 44 SO16 40

To target those likely to use a new transport service linking up locations on the West and East sides of the water with Southampton, interviews were carried out with households randomly selected from defined postcodes in the area around Southampton Water. The target area was split into 3 (East, Central, West) with 100 interviews achieved in each area, selected to give a broad distribution with respect to age, gender and employment status.

Connected Places

Age Range ALL Female Male 18-34 32% 31% 34% 35-44 25% 26% 25% 45-54 14% 14% 15% 55-64 11% 11% 12% 65+ 16% 18% 14%

Traveller Case Study:

From West Side

Connected Places

The journey by car into Southampton can be good on clear roads but when disrupted the journey time becomes very long for the distance covered. Some people report shifting work start time to 10am or turning down appointments before 10am. Locally, there are only trains from Totton to Southampton. Although the service is quick and good value, its lack of frequency (hourly) results in people choosing to drive. Also if you need to drive to the station you have to pay for parking. Buses are regular during the day but suffer from traffic issues on the A326 and do not run after 7pm. The cost of the bus was also cited as an issue – it’s just as cheap to drive and park.

Traveller Case Study:

From East Side

There are multiple driving routes into Southampton and parking is available (but can be expensive). Routes tend to be single lane roads and the flow is affected by the number of sets of traffic lights at

• junctions. Travel on M27 and M271 can be slow in the morning if

travelling around Southampton from the east. There are plenty of bus and train options including links to the airport and the bus services seem to be good albeit affected by road congestion. ‘Depending on the time of day it can take about an hour and 15 minutes just doing the last few miles, which is really infuriating.’ ‘The journey east-west through Southampton seems to be getting worse over the years. It’s always been quite bad at rush hour but now it seems to be quite bad a lot of the time’

Demand Modelling: Frequency, duration and mode

Understanding the typical journey purpose, frequency, mode and duration was a key target output, building a picture of the typical user needs.

Connected Places

Predictably, commuting was the most common reason for transport, with the car easily the most significant mode of transport for all journey types. These findings are consistent with the perceived view of transport in Southampton.

Shopping trips Visits to friends and family Commute to work Social occasions/evenings out Leisure/sport/hobbies Hospital, doctor or other appointments 14 mins 21 mins 22 mins 19 min 18 mins 18 mins

Frequency Mode Duration

Daily Weekly Monthly Car Bus Walk/Cycle Other

Demand Modelling: Overview

Understanding the likelihood of and reasons for travelers to use the service was a key target outcome. While a faster journey time was a strong motivator, just under half the participants would still consider using the service even if it delivered a slower overall journey time. Analysing the anticipated journey types that users would undertake on the service, taking into account journey time, was also an important objective. Assuming a similar journey time, respondents were broadly in favour of using such a service, though the lowest level of support was found in respect of commuting. Nonetheless, as the most frequent classification of journey, it is encouraging that 61% of respondents would be likely to use the service if the journey time were the same.

Connected Places

Faster journey Same journey time Slower journey Very likely Quite likely Not very likely Not at all likely

42% 32% 12% 10% 30% 39% 14% 18% 18% 31% 20% 31%

Leisure/sport/ hobbies Social

• ccasions/

evenings out Commute to work Visits to friends and family Shopping trips

95% 82% 61% 62% 70%

Same duration Faster Slower

Demand Modelling: Acceptance of Autonomous Control

A key offering of the concept is autonomous control, which could bring benefits to the level of service which could be

• ffered and the business case through lower operating costs.

Participants in both the focus groups and surveys were asked their likelihood of using the service if under autonomous control. Only 37% of respondents stated they would be less likely to use the service knowing it was autonomously controlled.

Connected Places

18% 20% 31%

‘Feels safer than an autonomous car – there’s less traffic and there’s probably a defined route’ ‘The key is whether the final destination is close to where the service terminates… another 20 minutes to the journey at either end then public transport becomes untenable and you are better off getting in the car.’ ‘It’s not the autonomy that concerns me, it would be the route that defines whether I’d use it or not’. ‘Any system that puts more vehicles

• n the roads in Southampton doesn’t

solve any problems. But this [water taxi] theoretically takes vehicles which would have otherwise gone into Southampton, off the road.’

37%

Less likely

56%

No difference

7%

More likely

Demand Modelling: Wider observations

Connected Places

• The appeal of using a car when several people travel together should be replicated by offering price incentives on

multi-ticket purchases.

• The timings of the service are of note, as stopping the service early is a barrier to current public transport offerings.

Running the service later and more frequently than other public transport schedules could increase uptake.

• The average price people selected was £5.97, although opinions were varied depending on the convenience the service
• brings. A willingness to pay study could be carried out once the specifics of the service e.g. docking points and whether

scheduled/on demand have been finalised.

• Security and safety need to be strictly enforced – with life jackets etc. but also potentially with a member of staff on

board, particularly for night time journeys or those which typically have low numbers of passengers travelling. There also needs to be a means of providing resilience and support during a breakdown.

• The most common type of trips undertaken are shopping trips, so docking points near main shopping destinations could

be appealing. Free parking could be a means to incentivising usage and to eliminate the current barrier to some other public transport modes too.

• Last mile provisions need to be thought out to provide the end-to-end journey convenience that a car currently
• provides. A ‘green’ option such as free bicycles or sectioned walking routes would help reinforce the environmentally-

friendly appeal of an electric boat.

Connected Places

Business Modelling

Business Case

To attract investment for both the development and

• peration of the service, a

sound business case needs to be demonstrated. This was conducted within the feasibility study. Building on the Rationale and Benefits identified through the focus groups and user surveys, a demand modelling analysis and commercial viability assessment were

• conducted. These were

built on estimates for

• perating costs and

expected revenues.

!"#$%&##' ()#&

Business Case

Rationale Commercial Viability Demand Benefits

Connected Places

Connected Places

Business Modelling: Demand modelling

For a proposed service to be successful, there must be a significant level of demand. In order to model the expected demand, the modelling within this feasibility study:

• was based on the current manned Hythe ferry travel patterns;
• accounted for the population and attractiveness of the neighbouring areas around the dock;
• incorporated perceptions on autonomy;
• accounted for an organic increase in demand as the service becomes more widely known and accepted.

This analysis could be replicated for other proposed services, given the necessary assumed data is available. However for the specific example of Southampton within this study, the highest demand for the service was predicted to be seen on routes with origin/ destination of Southampton or Northam with generally a lower demand for services connecting Netley and Hamble-le-Rice.

• 200

400 600 800 1,000 1,200 1,400

2021 2022 2023 2024 2025 2026 2027 2028 2029 2030

Annual Passenger Journeys (000's)

Demand forecasts for unmanned service based upon varying levels of service frequency and improved perception of service

30 Minute 20 Minute 15 Minute 10 Minute

Connected Places

Business Modelling: Cost and revenue modelling

Assessing the operating costs is critical to understanding the business case. As the vessel does not currently exist even in design format, a set of expected parameters were set to derive expected costs associated with a small electric autonomous fleet. The relative costs would vary by the nature and frequency

• f the service operated. As operating frequency was

identified as having a significant influence on the likelihood

• f use, understanding the variation of costs with frequency

was an important outcome in the overall feasibility assessment. Capital and maintenance costs of the ferries required to ensure the stipulated level of service were identified as the main driver of costs differences with changing frequency of service. Subsequent analysis was undertaken to identify the lines where most revenue is generated per boat.

• 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50

30 Minute Frequency 20 Minute Frequency 15 Minute Frequency 10 Minute Frequency Annual Operating Cost £m Ferry Purchase Costs Ferry Maintainance Costs Electricity Other

Business Modelling: Commercial viability

Assessing the overall commercial viability of the proposed service was a key output for this study in order to inform future investment in such a proposal. For a service to be commercially viable independently, revenue must exceed the costs

• f delivering that service. However situations

may arise where public bodies may provide financial aid or support for the provision of such services, if other mitigating factors provide a sufficiently strong weighting. From a commercial perspective, delivering a high level of service across all proposed docks is not efficient with costs significantly exceeding revenue generated. This is the case for the first five years post implementation, where demand for the service has not yet picked up. By 2030, the 10-minute frequency service is identified as the most solvent solution, where the increasing demand attributed to a higher level of service compensates the higher operating costs.

(£1,500) (£1,000) (£500) £0 £500 £1,000 £1,500 £2,000 £2,500 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Level of subsidy / Operating Surplus (£Ks)

LEVEL OF SUBSIDY (-) / PREMIUM FOR THE DIFFERENT LEVEL OS SERVICES

10 Minute Frequency 15 Minute Frequency 20 Minute Frequency 30 Minute Frequency Connected Places

Business Modelling: Economic benefits

The introduction of an autonomous water taxi service would support the reduction

• f traffic currently on the roads. This

could independently generate significant economic benefits relating to areas such as decongestion, accident reduction, emissions reductions, infrastructure benefits and noise level reductions. The magnitude of these perceived economic benefits have been assessed based on the assumed adoption rate of the service and consequential reduction in use of other transport modes.

Connected Places

Business Modelling: Economic challenges

Passengers will perceive time spend waiting, walking and interchanging for public transport double that time spent in the

• vehicle. Therefore time accessing piers and waiting for ferries will be

just as significant and painful as in vehicle time. A number of other factors will also drive modal choice which cannot be captured within the costs associated with journey time considerations and fares.

• Modal constant – factors relating to mode will influence choice. A car provides a more comfortable, secure journey

experience which will influence mode choice in addition to journey time considerations.

• Reliability – how journey time varies. Uncertainty around public transport can be much more discouraging than

uncertainty around car journey time variability.

• Individualisation of journeys – not all journeys are direct origin-to-destination trips. A car provides flexibility for a

passenger’s journey (e.g. school run on way to work) which can be a significant factor.

Connected Places

Business Modelling: Wider market for new water taxi services in UK

Connected Places

This feasibility study identified that ferry services can provide advantages over other transport modes, particularly in areas where coastal or rivers features provide severance which can easily be addressed by short water-based transport. In assessing the feasibility of any specific location to benefit from an autonomous water taxi service, a number of specific factors would have a significant impact on the outcomes. These factors would include:

• The size and distribution of the local population and their employment status and location
• The expected number of external visitors (tourism and business)
• The quality of any existing and competing public transport modes
• The comparative journey time savings (e.g. congestion or pedestrian/cycle severance such as lack of

crossing points such as bridges)

• The level of severance that rivers, lakes and estuaries currently cause in a given location.
• The impact of technology on competing modes

Connected Places

Safety

Regulation and safety overview

Connected Places

For the concept to be a viable long-term proposition the regulatory framework needs to be supportive of the technology. Marine autonomy is a relatively new topic, particularly so for passenger transport. To assess the feasibility of the concept means looking into the existing regulations surrounding the design, commissioning and operation (both on-board and

• ff-board) of such vessels.

At an International level, regulations are set by the International Maritime Organization (IMO). UK national waters come under the Maritime and Coastguard Agency (MCA). In addition, the viewpoints of and requirements from other stakeholders such as classification societies, insurance bodies and port operators are key. Understanding the fundamental safety of operating an autonomous passenger carrying water taxi was also a key

• bjective, achieved through a formal safety assessment

Connected Places

Safety: Through operation

The most significant regulatory barriers to the adoption of autonomous control for passenger-carrying operations lie within ensuring the safety of both immediate passengers and other seafarers not directly associated with the autonomous vessel. The issue of welfare, rather that the immediate control performance of the autonomous control, is of concern to organisations such as insurance bodies, the MCA and the port authorities in whose waters these vessels may operate. Most immediately, the Safety of Life at Sea (SOLAS) provides extensive requirements on many aspects of the design of vessels – notably, chapters IV (Radio Communications) and V (Safety of Navigation), though also some smaller impacts in other chapters. For passenger-specific applications, additional requirements apply such as a more rigorous survey schedule. Furthermore, specific requirements relating to passenger welfare, counting and recording passenger numbers, ensuring security of doors and hatches prior to launch and the general responsibilities for passenger welfare in the event of an emergency all apply but may not easily be addressable for an unmanned passenger vessel.

Connected Places

Safety: Through design

An initial review of the existing regulations suggest that an autonomous passenger-carrying vessel could not be permitted at this time under the existing regulations. In the design and classification process, many of the existing regulatory framework could still apply. Many of the existing requirements on the physical design and construction to ensure sea-worthiness and safety should apply equally whether autonomously controlled or not. Equally, the provision of safety systems on the vessels and any associated detection systems, and the requirements to adhere to the COLREGs. There are significant gaps surrounding how an autonomous control system would be deemed fit for purpose and operationally safe such that it could be classified and hence registered. For conventional vessels, the skills and qualifications required of the crew (as defined within the Standards for Training, Certification and Watchkeeping) and associated manning levels (as defined in various sections of SOLAS) would ensure this. These could not be met without crew on board, and there is currently no means to validate the correct control performance of an autonomous vessel against COLREGs.

Connected Places

Safety: Formal hazard assessment

To understand the feasibility from the perspective of safety requires more than a review of existing regulations and operational guidelines. Eight fundamental functions of the autonomous ship were defined, and using these a preliminary hazard assessment was used to inform specific considerations or recommendations appropriate for the concept.

Define Primary Functions Hazardous Failures Consequences Recommendations

The boat uses sensors to determine and process information of other objects and vessels within its sight-range The boat uses AIS data to track the position, type, speed and trajectory of other vessels fitted with and using the AIS system The boat uses GPS-based data to determine its own position within space The boat plans its initial path to reach the target destination in an efficient and safe manner, updating in real-time to avoid all obstacles in that path or coming into that path The route processor commands the propulsion and steering systems in order that the boat achieves the target path The boat docks and undocks with the dockside, with safety barriers to permit entry to / egress from the boat only when it is safe to do so The boat's progress and surroundings can be continuously monitored from a remote station and autonomous control overridden if needed The boat ensures the safety and protection

• f passengers within during the journey

Preliminary safety assessment

Connected Places

Understanding the safety hazards and barriers that could affect the feasibility of the concept was a key objective from the study. All of the identified hazards were rated and assessed. This highlighted specific areas of concern and high risk, some of which may be readily apparent and

• thers not.

Safety concerns centred on:

• performance shortcomings of the autonomous systems

leading to accidents

• safe management of passengers as they board and exit the

boat, with possible relative movement between the boat and the dockside

• ensuring the on-going safety of passengers while at sea -

both in terms of immediate passenger well-being and ensuring that any safety systems on board (life rafts / rings etc.) are present and correct A list of recommendations were made to address the hazards:

• Many of these are practical recommendations to ensure

robust autonomous control

• Several process related to ensure secure docking and

management of passengers boarding and exiting the boat

• A number relate to ensuring that communications to

and from the autonomous boats have sufficient redundancy measures due to the hazards associated with loss of communications and control in open water

Safety: Feasibility of the concept

Connected Places

• In a formal context, regulations are a barrier to the deployment of autonomous passenger-carrying vessels. Current

regulations stipulate minimum crew levels, which include trained radio-communicators and lookouts. Work is already being done by both IMO and MCA to understand if regulations can be adapted to support the deployment of autonomous passenger-carrying vessels

• Regulators do not currently have a way to determine the robustness of autonomously controlled vessels, either at launch
• r during their lifetime. Best practice guidelines exist, including the Maritime Autonomous Surface Ships UK Code of

Practice (Version 2), though are limited to vessels up to 24m

• The safety and welfare of passengers while at sea / afloat is a major concern for operators, regulators, insurers and legal
• bodies. Current guidelines or best practices do not yet consider provision to ensure the safety of passengers
• Provisions to manage the safe flow of passengers on and off the boat have not yet been addressed. Without any such

provisions or technologies, docking stations or the boat may still need to be manned

• Appropriate emergency response procedures where passengers are involved are not yet well defined
• Provisions for Port Authorities / Vessel Traffic Services to directly override or command an autonomous vessel within

their area of operation are not yet in place. A working solution could be communications to the vessel operators shore-based command centre

Connected Places

Project Outcomes

MANTA: Outcomes

The MANTA project considered the feasibility of an autonomous water taxi service operating in Southampton and on the

• Solent. In principle, the concept could be applied to other coastal locations provided suitable user demand is present or likely,

and hence a commercially viable service can be run. Key findings from the three areas considered may be summarised as:

Connected Places

20% 31%

There is some user appetite for such a service, though the convenience of use would be the major challenge:

• First mile/last mile journeys
• How seamless ticketing may be

between multiple travel elements

• Reliability and frequency of the

service Such a service would need some degree of subsidy, at least initially while user levels would be low

• A subsidy level just over £1.0m to provide a 10-minute frequency service in the first year or
• peration
• A subsidy level just over £0.4m to provide a 30-minute frequency service in the first year or
• peration

By 2030 the service could be profitable at either frequency level of operation, noting that the costs associated with a remote operating station and/or any land-based staff to support the autonomous service were not included in these calculations, and would therefore need to be explored further A number of a barriers exist in the safety and regulatory context:

• Existing regulations preclude autonomous passenger-carrying services, particularly in regard to watchkeeping, manning of radios and in

monitoring and maintaining passenger well-being

• Safe behavior in fault conditions needs addressing to identify the best approach
• Vessel management and interactions with port authorities needs development
• Safe management of docking and passenger entry/egress requires further work