Sparks

Problem Description

At a nighttime drone show, a formation of drones performs synchronized movement in the same plane. While moving simultaneously in the same direction, these drones occasionally throw colored sparks in all directions around them, these sparks providing the only visual clue regarding the position of the drones at a given moment.

For the purpose of proactive detection of unwanted accidents of the drone installation during the show, an automated system regularly takes photographs of the ongoing show and records the positions of the observed sparks.

Based on the positions of sparks detected at different moments, the goal is to detect the drones, respectively their number, and predict their position at a given moment in the future.

Dataset

The dataset.csv file presents the following columns in the header:

• Name: the name of the photograph
• Timestamp: the moment in time when the photograph containing the described spark was taken, in the format Jan 26, 2026, 12:00:00 AM
• X: the horizontal position of a spark, expressed in feet from the edge of the show area, as a real number
• Y: the vertical position of a spark, expressed in feet from the ground, as a real number

Requirements

Requirement 1

Determine the number of distinct photographs in the dataset and the number of drones in the formation.

Requirement 2

Determine the position of the drones at all moments for which there are photographs of emitted sparks.

Requirement 3

Determine the positions of the drones at the moment Jan 26, 2026, 12:03:20 AM, assuming they continue their movement unabated.

Evaluation

10 points are awarded by default, according to MEC regulations.

For requirement 1: 20 points for correctly determining the number of photographs and 20 points for correctly determining the number of drones in the formation.

For requirement 2: the percentage of correct answers multiplied by 40 points. A deviation of up to 0.7 feet from the expected value is accepted for either of the two coordinates.

For requirement 3: 10 points for the correct answer. A deviation of up to 0.7 feet from the expected value is accepted for either of the two coordinates.

For requirements 2 and 3, the order of drones represented in the file does not matter. For scoring, the evaluated values and expected values will be extracted, sorted in ascending order by X value, and compared 1 to 1.

Submission File

The submission file has the following header:

1subtaskID,Answer1,Answer2,Answer3

For requirement 1, exactly one row, as follows:

11,<distinct photographs>,<number of drones in formation>,<ignored value>

Attention! For requirement 1, the last value in the row is ignored during scoring, but must be present to not corrupt the CSV file format.

For requirement 2, ideally exactly (but no more than) <number of photographs> multiplied by <number of drones> rows, each of the form:

12,<timestamp>,<x-position>,<y-position>

representing the moment in time to which the record refers, followed by the horizontal and vertical position of one of the drones. A deviation of up to 0.7 feet from the expected value is accepted.

For requirement 3, one row for each drone, as follows:

13,"Jan 26, 2026, 12:03:20 AM",<x-position>,<y-position>

representing the moment in time given in the requirement, followed by the horizontal and vertical positions of each drone. A deviation of up to 0.7 feet from the expected value is accepted.