Although we have referred to it as feedback within the exoskeleton: the correct term is actually sensors. The word sensor already has the word sense in it and that is exactly what it is: an artificial sense. We have eyes, ears, and other senses. In machines, there are sensors as alternatives to these human senses. It is a piece of hardware that generates data. It detects changes in physical aspects, such as wind or light. The sensor then sends data on these changes in a readable format to a controller or processor.

Types of sensors

There are a lot of different types of sensors and they can be categorized in different ways. We will briefly cover the main categories, before diving into the ones we use at Project MARCH (and which are generally used in the exoskeleton world).

You can divide sensors into different categories based on what they measure. Examples of these categories are:

• Acoustic sensors: these measure sounds
• Flow or fluid velocity sensors: these measure fluids and how they behave
• Pressure sensors (such as a barometer): these measure air pressure
• Temperature sensors: these measure the temperature
• Current sensors: these measure the current
• Magnetic field sensors: these detect and measure magnetic fields, also called encoders

Another way to categorize sensors is by how they work. There are two main axes on which sensors can then be placed.

1. Active or passive sensors
2. Analog and digital sensors

Active sensors need to be powered by an external supply. Passive sensors do not need this, as they are built in such a way that they are powered by the thing they are measuring. An analog sensor differs from a digital sensor in many different aspects.

An analog sensor produces an analog signal. An analog signal is a continuous signal: it is active at all times. It also covers a continuous range of values of the physical aspect it measures. A temperature sensor, for instance, is an analog sensor. It tells you all the time which temperature it is and this temperature data has a continuous range from freezing to very hot and everything in between.

A digital sensor on the other hand sends a digital signal. A digital signal is transmitted at intervals. A digital signal only consists of 0’s and 1’s to represent information. Although this may seem quite useless, it is actually the most common signal used by computers. A DVD is an example of a digital signal. To make the difference between digital and analog even more clear, let’s refer back to the temperature sensor. A digital temperature sensor would for instance only tell you the temperature every fifteen minutes. And then it would only provide you with a 0 if it is freezing outside and a 1 if it is not.

Sensors in the Project MARCH exoskeleton

We have a few different types of sensors in the Project MARCH exoskeleton. Below we’ll explain for each of these sensors why we have them and how they work.

1. Temperature sensors

We use the heat sensors for safety so that our joints don’t heat up above a certain temperature. A too high temperature can mean something goes wrong within the joints, or our pilot may get burned by them. Since she cannot feel the temperature of the joints around her legs, we have heat sensors in each joint to measure their temperature. There are both digital and analog temperature sensors in the exoskeleton. The sensors tell us how high the temperature is. From the digital sensor, our master computer receives a signal which only tells us if the temperature is too high or not. We have a few different types of sensors in the Project MARCH exoskeleton. Below we’ll explain for each of these sensors why we have them and how they work.

In addition to the temperature sensors in the joints, we also have temperature sensors in the battery and master computer. Both the battery and master can start to seriously malfunction if they become too hot. If this happens, we shut down the exoskeleton so that it can have some time to cool off.

Something we learned along the way: if the weather gets above 30 degrees, your battery will heat up even more quickly! Here you see an impromptu cooling down session. Source: Project MARCH

2. Encoders

We use encoders in the exoskeleton to measure the angle of the joints. This information is essential for creating the gaits, as you have learned in the previous week. The way these encoders work is via magnetism. A magnet on the rotating joint passes by a sensing device, which can detect the changing magnetic forces. This allows us to know the position of the joint. However, there is another place where we use an encoder in the exoskeleton. In the video on the input device, we explained how there is a rotary button in the input device. Attached to this rotary button is a rotary sensor, which measures how far the rotary button has scrolled.

3. Pressure sensors

The pressure sensors are placed within the footsole of the shoe in the exoskeleton, therefore they are also called pressure soles. Pressure sensors measure exactly the same as what they are called, pressure. As an example, if the pilot stands on only his left leg a lot more pressure is applied on his left foot compared to his right. This information is sensed by the pressure soles. With this information we can get a much better estimation on how the exoskeleton is placed and how it is balanced.

4. Current sensors

Current sensors are crucial for the electrical engineers of project MARCH. The current sensors are part of the electronics and measure how much current is going to all the parts in the exoskeleton. With this information, several estimations can be made. It can be determined if a part of the exoskeleton is turned off if no current is sensed. Furthermore based on how much current is delivered an estimation can be made of how much power is used and how long we can still use the battery with that power usage.

5. Depth camera

Just like you use your eyes to look around you, the exoskeleton has cameras to look at its surroundings and make decisions based on that. However, a regular camera is not enough for that. To see for example how high a stair step is, it is needed to sense what the actual height is. You do this yourself unconsciously but the exoskeleton needs this information to calculate the gait it needs to perform. To get this depth information a depth camera is used. A depth camera works similarly to your eyes. It has two cameras placed closely together. The images they make are almost the same but not completely since they are not in the exact same spot. By comparing these images an estimation can be made of how far an object is.

The depth camera explained
(On the Computerphile YouTube channel)