The small arms race

In the wake of rising casualty rates in modern combat, the military is increasingly turning to robots for reconnaissance – and the race is now on to make them as small as possible.

Forget the traditional battlefield and troops advancing gingerly through fields of landmines. The front line of the future is robots – the smaller the better.

With conflicts increasingly taking place in populated areas and enclosed spaces, be they tower blocks or cave complexes, satellites and traditional reconnaissance are of little use. In laboratories around the world researchers are developing tiny reconnaissance robots for this new type of battle.

Modern electronics are enabling mind-boggling approaches, says Dr George Hayhurst, a senior research associate at the University of Lancaster’s Centre for Defence Studies: ‘The development of charge-coupled devices for use in miniaturised video cameras has helped a great deal. These CCDs are sufficiently small that some humourists have tried gluing them on to cockroaches.’

They have also been used on the recently reported ‘Robobug’. Based on a fly, this Flapping Wing Micro Aerial Vehicle (FWMAV) is being developed by Cranfield University researchers at the Royal Military College of Science. Research on it was initiated by Dr Rafal Zbikowski and funded by the Engineering and Physical Sciences Research Council and the Ministry of Defence. It has attracted a great deal of interest from the US government and military, including NASA’s Langley Research Centre, the US Air Force Research Laboratories and Army Research Office.

The military is showing special interest in Robobug because a flying robot like this could go anywhere carrying sensors such as synthetic eyes, ears and even a nose. The FWMAV could enter places inaccessible to small planes or helicopters and report back to a rescue team or military command post.

It can do this by using insect-style flight. Insects’ wings do not just flap up and down; instead, they trace a figure-of-eight pattern as they beat, flipping the entire wing upside down on the upstroke to create maximum lift and allowing the insect to hover. It is an efficient, low-powered method of staying airborne.

Zbikowski received £500,000 for three years’ work on two FWMAV projects. One finishes in July, while the other ends in 18 months’ time. By then he hopes to have a demonstration wing system that can flap realistically, and guidelines and mathematical models for further development.

As for the onboard sensors, the technology for those can come from all sorts of directions, says Hayhurst. Tiny CCDs are used by optical astronomers, who also developed sophisticated algorithms to produce high picture definition from low-quality raw data, which have obvious reconnaissance applications.

Hayhurst expects technology from many engineering disciplines to be brought together for micro aerial vehicle development, including techniques derived from nuclear weapons. ‘The Sandia National Labs, a subsidiary of Lockheed Martin, has an interesting sideline on micro machines. This technology has been used in the miniaturisation of the optical fire-chain mechanisms for refurbished nuclear bombs.

But it can also be used for other devices in miniaturised unmanned aerial vehicles (UAVs),’ he says.

Sandia researcher Doug Adkins has been working on these micro machines. In 1996 his research team unveiled the Mini Autonomous Robot Vehicle, a 16cm3 robot with all the power, sensors, computers and controls on board. Since then further miniaturisation work has brought about MARV Junior. Using rapid prototyping methods to build the mini-vehicle’s frame layer by layer, Adkins’ team has created a robot that measures just 4cm3 and weighs less than 1oz. Within it are cavities for batteries, electronics, axles, tiny motors, switches and other parts.

Adkins had to rethink how the smaller device would move around. The first MARV had standard wheels but, as mobility was limited by the smallness of the wheels, tank tracks were used on its successor. He says: ‘I thought of how tank tracks can manoeuvre over large objects and realised the mini-robots could benefit from the same type of wheels. They could initially run for about 10 minutes with them, but now they can operate for 50 minutes using silver oxide watch batteries.’ With tracks the robot can also move on carpet — particularly important for potential house-to-house reconnaissance.

The Sandia micro robot has already manoeuvred its way, at about 5cm a minute, through an obstacle course of coins. But to control that movement with the even smaller MARV Junior a new approach to the electronics had to be taken. Adkins’ team mate Edward Heller used commercially available unpackaged electronics parts (without plastic casings) to overcome this.

Sandia’s Compound Semiconductor Research Laboratory then assembled the electronics on to a simple multi-chip module on a glass substrate. Heller says: ‘Previous small robots consisted of packaged electronic parts that were more bulky and took up valuable space. By eliminating the packaging we reduced the size of the robots’ electronics considerably.’

Marv Junior is powered by three watch batteries. It consists of an 8kb ROM processor, temperature sensor and two motors that drive the wheels. A miniature camera, microphone, communication device and chemical micro-sensor for detecting chemical weapons are among the enhancements being considered.

With these improvements the micro robots could scramble through pipes or prowl around buildings looking for chemical clouds or even human movement. They could relay information to a manned station and communicate with each other via infrared or radio wireless capabilities. They will also be able to work together in swarms, like insects. Eventually they may be capable of performing difficult tasks that are already carried out by much larger robots today, such as locating and disabling landmines or detecting chemical and biological weapons.

Flying spies

So far the only interest Sandia has had for its devices has come from toy manufacturers. But while these robots continue to be laboratory curiosities, firms such as Honeywell and Micro Craft have won contracts for robots the size of a portable television.

The US Defence Advanced Research Project Agency is spending $33m (£23m) on two programmes for these flying spies. The Organic Air Vehicle and Micro Air Vehicle Advanced Technology Concept Demonstration programmes will run for two and four years respectively. The plan is to create a robot vehicle that will be carried to the battlefield by a soldier.

The OAV programme will test autonomous prototypes that will be able to operate in adverse weather conditions for an hour and with a range of 10km; trials are expected to begin in December. The MAC-ATCD project will test similar vehicles with the ability to take off and land vertically; around 350 will be made, some of which will be used in ‘in-field’ trials in military operations. One take-off-and-land helicopter already under trial is the Hummingbird.

That and the OAV form part of the US military’s Future Combat Systems programme. This long-sighted project envisages a radically changed army, navy and air force with unmanned autonomous land, sea and air vehicles, of various sizes, a key aspect. Says defence expert Dr John Alexander, a consultant to the US government on non-lethal weapon strategy, a former commander of Green Beret A-Teams in Vietnam and director of the US Army’s Advanced Systems Concepts Office: ‘Future Combat System is the programme to take a present-day combat force, such as a brigade, and achieve double the effectiveness with half the people and a quarter of the weight of current weapon systems.’

This philosophy stems from a major financial decision the US armed forces had to make many years ago. Should it spend its money on weapon systems that had huge conventional destructive capabilities with few resources focused on reconnaissance, a policy of guess where the enemy is and flatten it? Or should it go for excellent reconnaissance assets that could identify individual targets for precision weapons?

The military chose the second, and the world saw the policy at work in Afghanistan: unmanned Predator drones flying at 26,000ft, controlled remotely, watching for Al-Queda members, to guide troops and bombers in. It was sufficiently impressive to persuade the UK’s MoD to spend £19m on the next phase of development of UAVs. BAE Systems, Lockheed Martin, Northrop Grumman and Thales are competing to provide unmanned aircraft under what could eventually be a £400m contract. Under the current phase each will provide details of their system’s capabilities, as a result of which the four will be reduced to a shortlist of two.

This week the European Aeronautic Defence and Space company launched its own range of UAVs at the Singapore Air Show. The company has developed micro drones and Predator-like tactical craft. It is also working with US defence firm Northrop Grumman on a European version of the US military’s high-altitude long-endurance Globalhawk.

All in all, military surveillance looks certain to become increasingly robotic, but what level of autonomy will the robots have? Alexander is under no illusion: ‘Robots that can shoot and kill can be very small. For urban locations the robots will have to be able to creep, crawl, enter rooms and use their sensors to transmit information back to identify terrorists.

The only real question is: who makes the kill decision?’