This article was first published in the Ricardo Quarterly Review. It is used here with permission.
Now that most of society has embraced the idea that the world's resources are limited, the appeal of autonomous systems in vehicles which help the driver to optimise driving style for better fuel economy is much greater. Car ownership, personal mobility and, especially, driving skills are all emotive subjects - yet at the same time it's well understood in engineering circles that driving style is one of the main causes of excessive fuel consumption. Safety specialists agree, too, that it is often the driver rather than the vehicle that is the weak link in the chain.
Based on a Ford Escape hybrid, the Ricardo Sentience research vehicle addresses those problems without taking control of the vehicle away from the driver. ‘Sentience’ literally means 'consciousness' and by providing the vehicle with a degree of intelligence, information on the topography of the road ahead and real-time traffic information, it can function more efficiently and safely. Sentience exploits the benefits of both advanced hybrid technologies and intelligent transport systems to the full, reducing fuel consumption in track-based testing from anything between 5 and 24 per cent.
The Escape is a full hybrid which can operate in several modes. It can run under electric-only power, using the power of just the petrol engine, or operate in mixed mode where both the petrol engine and electric motor are brought into play. In regenerative braking mode, kinetic energy is recovered by using the electric motor as a generator to charge the battery when the vehicle is slowing.
A comparison of test data from the Sentience vehicle - operating its Enhanced Acceleration/Deceleration (EAD) strategy - with average vehicle use across mainland UK roads, suggests a national fuel saving of up to 14 per cent may be possible, the equivalent of between 1.2 and 2.9 million barrels of oil per year. Yet despite these impressive predictions, the cost of implementation is minimal and has been achieved by clever combinations of existing technologies already in the marketplace and which are installed on many existing vehicles.
"Sentience has been a unique collaboration between three industries which would not normally work together," explains project director, Tom Robinson of Ricardo. Other project partners are Jaguar Land Rover, TRL, Orange Business Services, the UK mapping agency Ordnance Survey, and innovITS, the UK centre of excellence for intelligent transport systems and sustainable mobility, which part-funded the project.
EU targets demand intelligent systems
Clearly, there is a pressing need to improve both CO2 emissions and safety on the roads; nevertheless, the project team believes that EU targets are unlikely to be achieved through conventional means. Current safety measures are expected to deliver a cumulative reduction in fatalities of 40 per cent by 2020 but the EU target is 75 per cent. CO2 emissions reduction has already fallen well behind targets, as the 140 g/km average by 2008 agreed by ACEA has not been met. Instead, an EU-wide fleet average of 120/130 g/km will be phased in between 2012 and 2015, with heavy fines in store for those manufacturers who fail to comply: the EU's medium term target is 95 g/km by 2020.
The steady increase in the intelligence of vehicles is one way of improving both CO2 emissions and safety. Driver information systems can assess driving technique, provide real-time advice to drivers and supply 'green' routing directions for optimal fuel consumption. As technology progresses, semi-autonomous control will help to overcome the inadequacies of the driver and will provide longitudinal control over acceleration and deceleration. Further down the line, fully autonomous control may provide intelligent traffic management with both longitudinal control and lateral control as well as vehicle-to-vehicle communications.
There were three phases to the Sentience program. The first lasted six months and encompassed simulation and specification. The following 9 months were spent on implementation and assessment, while the third phase was devoted to demonstrating the technology. The initial step was to create and validate a model of the Escape to assess its performance. This was accomplished on a variety of routes, such as a flat 20 km route and a second 20 km route containing sections of flat, uphill and downhill sections, covered at different speeds with air conditioning on and off. Other routes were also simulated with both head and tail winds to provide an extremely comprehensive picture.
Opportunities for energy saving
The next steps were to assess the opportunities for saving energy using regenerative braking by exploiting the vehicle's ‘extended horizon’, through the use of telematics and knowledge of the road ahead.
Ultimately, the amount of energy captured is limited by the battery capacity and the size of the electrical machines generating the electricity during regenerative braking. Yet the vehicle can also analyse which are the best points during the route to use electrical power, given that it knows what the traffic will be like as well as knowing the detailed topography of the roads it will soon encounter. The potential for the car's ability to look ahead was assessed in detail during the simulation phase.
Finally, the car was modified with the components it needed. These included a bespoke wiring harness, a modified air conditioning unit, a custom cruise control box, Ricardo's rCube-based Sentience control unit, a wireless router and receiver for in-vehicle communications and a CAN data logger. All of these components are out of sight, leaving the Escape looking perfectly standard except for the mobile phone sitting in its windscreen-mounted cradle and the installation of a small GPS receiver.
On the Road
So how does all this translate to real world driving? It has often been said that the essence of good driving is anticipation. And that is what Sentience is really all about, anticipating what lies on the route ahead then planning the use of the car's energy resources as efficiently as possible.
Most drivers will understand that accelerating hard from a green traffic light to the next red light and then braking hard is an inefficient way to drive and wastes fuel.
However, most drivers probably do not have the technical knowledge to modulate the engine load to best effect on different gradients and conditions for best fuel economy. Neither has any driver the ability to know that a few hundred metres ahead around the next blind bend, there is a traffic queue forming and that the best course of action would be to start slowing gradually to save fuel.
Reading the road ahead
Sentience can do all of these things and more by reading the road and traffic conditions ahead and adopting the most efficient strategy for the hybrid powertrain accordingly. To do this, it focuses on three areas: Enhanced Acceleration/Deceleration (EAD), Optimised Engine Load (OEL) and Enhanced Air-Conditioning (EAC).
EAD is a form of adaptive cruise control and makes sure the vehicle adheres to speed limits while driving in the most economical way.
"EAD reduces the speed variability which occurs with most drivers and which impacts on fuel consumption. But it doesn't bring the vehicle to a complete halt - that still remains the responsibility of the driver," explains Robinson.
OEL manages the hybrid systems for optimum use of either the electric or combustion engine by drawing on advanced knowledge of the route ahead. The powertrain controller was modified by Ford to allow integration with the Sentience system, and the Ricardo rCube rapid prototyping platform is used to execute the control strategies as well as interfacing with the vehicle systems.
"OEL looks for opportunities beyond where you are now to provide an overall gain," says Robinson. "It decides which energy source should be used when, by considering the whole route."
EAC manages the air conditioning ahead of temporary stops. The air conditioning compressor is driven by the engine, so when the engine is stopped (for example when the hybrid is running in electric-only mode, or at a halt) there would normally be no air conditioning. But by linking the air conditioning control system to that of the hybrid drive, again overseen by Ricardo's rCube, knowledge of when the combustion engine may be stopped will allow the system to plan an appropriate cooling strategy. This control system keeps the cabin temperature within a narrow range while still allowing the gasoline engine to switch off and hence minimising the time that the engine is run solely to drive the air conditioning. Ricardo has applied for a patent on the EAC control system used, which is applicable to all types of hybrid vehicles including the "start-stop" systems which are becoming increasingly common.
Intelligence to look ahead
Sentience's intelligence comes from a combination of advanced mapping data, internet-enabled mobile communications and GPS. The mapping information is based on what Ordnance Survey calls the OS Mastermap Integrated Transport Network layer. This is an advanced form of mapping that includes much more than the usual topographical information, and provides details of gradients, curves, speed limits and features that may affect speed, such as crossings, schools, traffic calming measures and traffic lights.
The project team call this ‘electronic horizon' information and it is created by a combination of route driving, data logging and manual data entry. Road routes were prepared near Ricardo, TRL and Orange sites, in addition to the six routes on the TRL proving ground.
"The data allows the vehicle to look ahead for some distance," explains Robinson. "1 km is good but 2.5 km is the optimum. The dynamic programme then works out the best strategy for the next part of the route."
But there's more to the data side of things than the extremely enhanced mapping: this is where mobile communications provider Orange plays to its strengths. Orange provides the HMI (human machine interface) in the form of a Nokia N95 mobile phone configured with bespoke software or a so-called 'widget'. Once the widget has been launched, the driver can enter the desired destination, get topological details of the route, see the state of charge of the battery and adjust the enhanced air-conditioning functions. The user can configure various functions and view pop-ups which display new topological features that that may affect Sentience.
The phone provides the real-time link with the outside world needed for providing live traffic information, for example, and the prevailing speed limit in force at any given point. Ricardo and Orange worked closely together to forge a link between the HMI and Ricardo's rCube, the rapid prototyping controller which supervises the important Sentience EAD functions. The phone is also connected to a GPS device enabling it to track the location of the car.
So, what can drivers expect to happen when they try the Sentience vehicle operating the EAD strategy for the first time? Well, the first thing is that there is no real need to adapt to anything particularly new because Sentience integrates neatly with the Escape's native cruise control system. All the driver has to do is switch on the cruise function and then Sentience takes over. But instead of simply maintaining a set speed like a cruise control system, Sentience will slow for junctions, corners, and traffic lights so that the driver steers throughout and brakes when necessary. In this way despite its high level of automation, Sentience still requires that the driver is the final authority for decisions regarding braking and emergency manoeuvres, such as stop signs, red traffic lights or for the unexpected such as if a pedestrian steps out.
Integrating with vision-based systems
The system could, in the future, integrate with vision-based systems which can identify road signs and pedestrians, for example, as well as with autonomous braking systems. These are already entering the market and can stop the car in an emergency if they see an obstacle suddenly appearing. There are also opportunities to link the system with future vehicle-to-vehicle (V2V) communications devices, which are currently under development by a consortium of OEMs and suppliers. 'V2V' uses wireless networking to allow intelligent interaction between vehicles for the purpose of traffic control and active accident avoidance.
For the moment, however, the main emphasis has been on EAD because it is this aspect of driving which can best be improved by the use of Sentience and show a substantial potential for reducing CO2. All types of roads, from motorways to minor roads, have been catered for in both the simulation and modelling work and in the physical testing carried out at TRL on test routes.
Test results show major potential
In all, the Sentience vehicle completed many hundreds of laps across six different routes at TR1’s proving ground and 612 km on the road routes. EAD succeeded in strictly maintaining the speed limits and, as expected, the variability of speed was less than when under driver control. In a production implementation it is likely that a Sentience-enabled vehicle would provide a user-configurable trade-off between fuel saving and journey time, as well as a dynamic routing capability.
With EAD enabled, fuel efficiency was improved by between 5.3 and 23.9 per cent in a variety of simulated conditions involving high and low traffic density and a mixture of urban roads and motorways. Road trials, which took place in the evenings, returned a fuel saving of more than 5 per cent.
Climatic dynamometer testing of the Sentience vehicle was undertaken by Ricardo to evaluate performance with Enhanced Air Conditioning (EAC) enabled. Under temperatures typical of UK summer weather conditions, fuel consumption over the urban portion of the NEDC cycle was reduced by more than 9 per cent in all tests, in comparison with operation of the vehicle's standard air conditioning system. Further dynamometer-based testing is required to conclusively establish OEUs impact on fuel consumption, but the project team predicts savings of between 4 and 9 per cent.
Low cost of implementation
But perhaps one of the great strengths of Sentience is the low cost of implementation.
"In the future, the EAD, OEL and EAC functions of Sentience could simply be extensions of the navigation, powertrain control and climate control systems," says Robinson. "In a vehicle already equipped with a phone and satellite navigation, no additional hardware would be necessary. But that doesn't mean Sentience would be restricted to vehicles with a comparatively high specification. Assuming high volumes, a standalone system could be incorporated for around $40."
While control and integration software would be needed, any additional processing and storage capacity would be modest and relatively easy to integrate into the architecture of most vehicles. The system relies on the availability of high-resolution mapping data but that is already becoming available and is not expected to pose any problems.
On that basis, the project team expects that, technically speaking, the entire Sentience system could be production-ready in just three to four years, with individual sub-systems such as EAC air conditioning control available now for immediate integration with future production programmes. Considerable interest has already been expressed by many OEMs and the Sentience vehicle is currently being demonstrated to potential customers around the globe.