Month: December 2018

We’re Hiring – Jobs at DriveTrust

We have received an investment and currently are looking for people who are interested in working in a start-up.
The areas you could work on are:

    • Embedded Firmware
    • Computer Vision
    • Hardware
    • Linux Applications
    • Mobile Applications

In case you are interested please click the button below and submit an application!

Hardware Engineer

Responsibilities

  1. Taking part in PCB design, validation and testing
  2. Taking part in hands-on prototyping
  3. Debugging activities
  4. Procurement of selected components and prototypes.

Required Skills

  • The engineer is comfortable with using CAD tools for multi-layer PCB layout. e.g. Altium Designer, CircuitStudio or CircuitMaker
  • Using hardware debugging and analysis tools such as oscilloscope, multimeter, signal generator, logic analyzers
  • Ability to read datasheets
  • Reading and designing electronics schematics

Optional Skills

  • Knowledge of real-time systems
  • 3D modeling skills for illustration and 3D printing
  • JTAG debugging
  • Experience with memory layout and camera
  • Experience with industrial design and enclosure design

Firmware Engineer

Developing drivers and integration software between hardware devices and the operating system. The devices include communication modules, a camera, sensors, a GNSS module and potentially external coprocessors.

Responsibilities

  1. Developing and porting Linux device drivers, such as ISP driver and communication module drivers.
  2. Compiling a custom kernel
  3. Porting open-source device drivers
  4. Ensuring the Linux OS can communicate with devices in an optimal way

Required Skills

  • Fluent in C
  • Comfortable with Linux environment
  • Understanding of compilation process, computer architecture
  • Understanding of interfaces such as I2C, SPI, MIPI-CSI and similar
  • Ability to read datasheets

Optional Skills

  • Experience with ISP drivers
  • Experience with Gstreamer and V4L2 pipeline
  • Experience with OpenCL
  • Fluent in C++
  • Experience with ARM architecture and ARM Assembly
  • Experience with memory layout
  • Experience with camera interfaces
  • Knowledge about communication protocols and modules e.g. Bluetooth 4.x/5, LTE, WiFi

Application Engineer

Developing the DriveTrust application for the DriveTrust device. In terms of the application environment, the DriveTrust device can be compared to Raspberry Pi.

Responsibilities

  1. Get data from sensors and processes/threads
  2. Combine the data into metrics  e.g. three strong acceleration events in two minutes = aggressive driver
  3. Send the data to cloud over the mobile network (mobile network module configuration might be required)
  4. Second Prio: working with Wifi, Bluetooth, V2X communication

Required Skills

  • Speaks English
  • Linux native
  • Proficient in Python

Optional Skills

  • C/C++
  • Android / iOS development

Computer Vision and Data Engineer

We will have a camera and would like to detect objects in real time from the video. Objects we would like to detect are cars, pedestrians and street signs. The MVP is a camera which can accurately detect street signs and traffic lights. On top of that depending on complexity we would like to have a lane departure warning system (LDWS) and a front collision warning system (FCWS)

Responsibilities

  1. Develop / Train a neural network for detection of cars, pedestrians, traffic lights, LDWS and FCWS
  2. Optimizing the network for the constrained hardware which is used
  3. Optional: combining object detection with sensor data to improve precision, e.g. getting speed from GNSS to improve FCWS
  4. Optional: Filter Sensor Data from GNSS / accelerometer to detect certain events

Required Skills

  • Speaks English
  • OpenCV
  • Experience with object/image detection frameworks (Tensorflow, Caffe or similar)

Project Assistant

Responsibilities

  1. Administrative tasks:
    1. Procurement of hardware parts/travel booking/  basic bookkeeping
    2.  Interface to accounting with University Rennes 1
    3. Support with French legal requirements (like starting a company/opening bank account)
  2. Communication:
    1. With partners and local in Rennes(organizing meeting/events etc.)
    2. With potential customers/businesses who are interested in us
  3. Content Creation / Online Research (optional, second priority):
    1. Maintain a blog, PR, basic online marketing
    2. Business Research: Funding / Investment programs, Potential customers etc.

Required Skills

  • Speaks English and French (Preferably C1 for both languages)

Driver Evaluation

We distinguish between two categories of measurements. The “classic measurements”  corresponds to driver evaluation based on already existing technologies and which can be accomplished with GNSS and accelerometer. The new metrics developed by DriveTrust rely on the camera on DriveTrust’s real-time object detection capabilities.

Classic Driving Behaviour Evaluation

Those metrics are used by other devices already existing in the market. DriveTrust collects the following data and combines them into profiles:

  1.  Generic driving behaviour, aka the “generic profile” which consists of:
    • total distance driven and average distance driven per month
    • nighttime hours driven
    • hours driven in bad weather conditions
  2. Aggressive driving which we combine to the “aggressiveness profile” is built from the
    • number of harsh acceleration and braking events
    • strong acceleration in curve events

Innovative / New Driving Evaluation Metrics

DriveTrust is developing new driving behaviour evaluation metrics. For this purpose, we work closely with our car insurance partners. There are three main metrics categories:

  1. Reaction Time: we measure reaction time to traffic signs and traffic lights, which gives a “reaction profile” for each driver
  2. Distance to other cars: the combination of speed data from GPS and the distance estimate from the camera provides insight into the “tailgating profile” of the driver.
  3. Lane Departure: we track if the driver slowly drifts into the left or right side of the road. Hence, we generate a “drifting profile” for each driver
Example mapping of a driver behaviour based on profiles

The combination of the reaction, tailgating and drifting profiles give the possibility to evaluate drivers better than with already existing technologies.