POD Technology

Aim Technologies

Specialising in mobile autonomy, navigation and perception, Importantly, its systems require no connectivity – the autonomous vehicles map and “learn” their environments for themselves. Having learned the route, the system constantly compares what it sees to what it expects to see, which enables it to calculate its precise location (to within two or three centimetres) and also recognise hazards such as pedestrians.

Aims Background and experience with aviation, OEM and motorsport, and sensors includes advanced radar, high performance computing, automotive systems and vehicle operations. For autonomous vehicles, its new AI system consists of modular sensing and processor units that can be tailored to individual vehicle and operating requirements.

Aim AI System uses complementary sensor technologies including radar and cameras, combined with high-performance processing units that run our own proprietary algorithms. This enables the full FMS System we have to provide an all-round situational awareness capability and vehicle control function, that can operate effectively in the most challenging conditions.

Environmental Advantages

The Autonomous POD system has zero emissions – it is fully electric and is the most environmentally friendly way of transporting people over relatively short distances. This is because its utilisation is optimal. Whereas a bus has to run along a given route whether it has one passenger or 50, the POD is an on-demand service – only vehicles that have been called by a user will move. The vehicles will only ever carry people or goods: they will never circulate carrying just air.

Safety

The safety of the vehicles has been proven over millions of kilometres at Heathrow Terminal 5. With regard to the safety of other road users, the POD has taken part in three UK government trials to ensure, as far as is humanly possible, that the vehicles will not endanger pedestrians or cyclists. While it cannot be proven that the POD can never have an accident, it has been demonstrated that the risk of an accident is lower than with any other form of wheeled transport.

The Guideway Application

If a user already has a transport infrastructure in place, or intends to install a dedicated infrastructure (e.g. at an airport where the volume of traffic justifies a separate roadway for the vehicles), the POD can use guide-way sensors along the route to control the pods. This means the vehicles are lower cost, as they do not need their own navigation system, but the infrastructure cost is higher.

The Shared-Space Application

For many applications, such as from a train station to a hotel, it is more cost-effective to use the vehicles on existing roads or paths where the pods need to safely interact with other road users. In this case, the vehicles “learn” their environment before going into public use. The learning process is rigorously controlled by the POD:

STEP 1

The vehicles are under the manual control of an operator, who guides the vehicle along a given route. The vehicle sensors build a highly sophisticated 3D map of that route, noting every feature – e.g. lamp-posts, kerbs, white lines in the road.

STEP 2

The vehicle then goes along the route autonomously at very low speed (under 5 km/h). At each road marking or piece of street furniture it encounters, it stops briefly to compare what is sees to what its map says it ought to see. When it is confident that the two images match up, it moves to the next item. At this stage a controller is in the vehicle in case of any problems.

STEP 3

It repeats the route numerous times at progressively higher speed, until it is sure that everything its encounters is exactly as its map predicted. It repeats this in different weather conditions (sunlight, cloud, rain, night-time etc.) Again, a controller is in the vehicle.

STEP 4

Once the vehicle is confident about the route, the POD introduces its own staff to the route as pedestrians and cyclists to ensure that the vehicle correctly identifies other road users in that particular environment (e.g. there are no reflections on the route that could stop the sensors identifying another road user).